Accepted Manuscript

How well do women do when it comes to financial literacy? Proposition

of an indicator and analysis of gender differences

Ani Caroline Grigion Potrich, Kelmara Mendes Vieira, Guilherme Kirch

PII: S2214-6350(17)30102-8
DOI: https://doi.org/10.1016/j.jbef.2017.12.005
Reference: JBEF 126

To appear in: Journal of Behavioral and Experimental Finance

Received date : 16 November 2017

Accepted date : 16 December 2017

Please cite this article as: Potrich A.C.G., Vieira K.M., Kirch G., How well do women do when it
comes to financial literacy? Proposition of an indicator and analysis of gender differences. Journal
of Behavioral and Experimental Finance (2017), https://doi.org/10.1016/j.jbef.2017.12.005

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
Cover Letter

How well do women do when it comes to financial literacy? Proposition of

an indicator and analysis of gender differences

ANI CAROLINE GRIGION POTRICH, KELMARA MENDES VIEIRA AND

GUILHERME KIRCH

Ani Caroline Grigion Potrich ([email protected]) is a Professor at the Graduation

Program in Business Administration, Federal University of Santa Catarina, Brazil; Kelmara

Mendes Vieira ([email protected]) is a Professor at the Post Graduation Program in

Management of Public Organizations, Federal University of Santa Maria, Brazil; and

Guilherme Kirch ([email protected]) is a Professor at the Post Graduation Program

in Business Administration, Federal University of Rio Grande do Sul, Brazil.

Corresponding author:

Kelmara Mendes Vieira

[email protected]

Postal Address: Roraima 1000, 74C, 4212, Santa Maria, RS, Brazil. Zip Code 97105-900

Tel.: +55 55 3220-9316.

*Manuscript

How well do women do when it comes to financial literacy? Proposition of an indicator

and analysis of gender differences

Abstract

National strategies aiming to reduce gender inequalities should focus on critical areas, such as

financial literacy. However, a method should be established to measure individual financial

literacy. We thus develop an indicator to assess financial literacy level and we analyze gender

differences. It is possible to identify whether individuals have a low or a high level of

financial literacy, as well as to draw a profile of the most vulnerable group. We carried out a

survey of 2,485 individuals in Brazil and the main results showed most individuals have a

low level of financial literacy across both genders. A significant relationship between

financial literacy and gender is observed; the proportion of men is higher among those with a

high level of financial literacy. However, the conclusions suggest that greater efforts should

be made to reach women, particularly those who are single and have lower levels of

education and income.

Keywords: Financial literacy; Indicator; Gender; National strategies.

1. Introduction

Recognition of women’s rights in different dimensions of life is reflected in the

historical approach to gender in treaties, agreements, and international conventions, as well as

in national policies aimed at ensuring these rights (IBGE, 2014). Adoption of national

strategies aimed at reducing gender inequalities demands an information system supporting

the formulation and monitoring of public actions that reflect critical areas within gender

issues. Specifically, one of issue is discuss by Bucher-Koenen, Lusardi, Alessie, and Van
 2

Rooij (2017), which argue women live longer than men and are likely to spend time in

widowhood.

Guided by this framework, governments of both developed and emerging countries are

concerned about financial literacy, as this is one of the factors contributing to uninformed

financial decisions, which in turn have substantial negative effects on society and on

individual countries (Gerardi, Goette, & Meier, 2010). In this context, governments are

interested in finding effective approaches to improving the financial literacy of their

populations by creating or improving national strategies, to provide learning opportunities at

various educational levels and for the most vulnerable groups (Atkinson & Messy, 2012).

However, prior to creation of such strategies, an indicator is required to determine

individuals’ financial literacy level, as well as their main needs, to develop effective

programs.

The literature already indicates that socioeconomic and demographic factors influence

the level of financial literacy (Chen & Volpe, 1998; Lusardi & Mitchell, 2011; Atkinson &

Messy, 2012; Agarwalla, Barua, Jacob, & Varma, 2015; Bucher-Koenen, Lusardi, Alessie, &

Van Rooij, 2017). Gender is one of the main variables surveyed. Evidence shows that women

have lower financial literacy levels than men and that they have more difficulty than men in

performing financial calculations (Atkinson & Messy, 2012; Mottola, 2013; Agarwalla,

Barua, Jacob, & Varma, 2015). Women also do not seem to master basic financial concepts

and they show a lower level of financial knowledge, which makes it difficult to take

responsible financial decisions (Sekita, 2011; Bucher-Koenen, Lusardi, Alessie, & Van

Rooij, 2017).

This situation becomes more worrying as women gain ground in the labor market and

participate in more consumer decisions, income management, and debt decisions. Further,

women are now often heads of households. Between 2000 and 2010, for example, the
 3

percentage of Brazilian women responsible for households increased from 22.2% to 37.3%,

according to the Brazilian Institute of Geography and Statistics Census of 2010 (IBGE,

2014). Already in 2015, this number reached 40% (IPEA, 2017); this demonstrates the

importance of women to the country of this study, which are increasingly in charge of

Brazilian families.

However, despite the importance of this issue, there is still no consensus on the best

tools to use for the definition and measurement of financial literacy. One of the most widely

used definitions was developed by the Organization for Economic Co-operation and

Development (OECD), which defines financial literacy as “a combination of awareness,

knowledge, skill, attitude, and behavior required to take financial decisions and ultimately

achieve individual financial well-being” (OECD, 2013). However, a widely agreed model

does not yet exist, since the literature argues that financial literacy is a multidimensional

concept for which a single construct would be insufficient to cover all dimensions involved.

Therefore, there is a need to develop and validate models that examine different

measurements and their interrelationships simultaneously.

The objective here is to fill this gap by developing an indicator to assess financial

literacy and to analyze gender differences based on the method proposed by Potrich, Vieira,

and Kirch (2015). These authors propose a method for development of a financial literacy

indicator; however, they find that their measure of financial attitude and financial behavior

proves unstable. Therefore, the present paper advances this work to propose a measure more

robust for the Brazilian context, via inclusion of additional dimensions.

We apply a methodology that integrates the three main pillars of financial literacy:

financial knowledge, financial attitude, and financial behavior (OECD, 2013). We seek

specifically to develop and validate an instrument to assess the level of financial literacy; to

implement a methodology for calculation of the financial literacy indicator; to analyze the
 4

differences in financial literacy levels between men and women; to identify the

socioeconomic and demographic profile of the most vulnerable group; and to suggest steps

that might be taken to improve the financial literacy of individuals, such as proposals for

public policies focused on the most vulnerable group.

This study is justified by the need to develop a model that gives insight into the

financial literacy level of Brazilians. Further, the literature provides evidence that the

financial literacy of women is lower than that of men, but it does not analyze within a

restricted sample of female individuals how this level of financial literacy varies with their

demographic profiles. The proposed indicator will be useful for economic agents to assess

financial literacy levels and to assemble adequate strategies and financial products for their

customers according to their profiles. From the government point of view, such an indicator

can give insight, for example, about the most vulnerable groups, allowing focused action

toward improving the financial literacy level of these groups.

2. Background

(a) Financial literacy

The modern economy increasingly requires more complex and sometimes confusing

choices from the consumer. An essential skill then emerges: financial literacy. However,

there are some gaps in the key concepts involved. The first is the fact that the term financial

literacy is often used as a synonym for financial education or financial knowledge. These two

constructs are, however, conceptually different, and using them as synonyms can lead to

problems, because financial literacy goes beyond financial education. Huston (2010) argues

that financial literacy has two dimensions: understanding, which is personal financial literacy
 5

or financial education, and its use, that is, the application of such knowledge in personal

financial management.

According to Savoia, Saito, and Santana (2007), financial literacy is the process of

acquiring knowledge and skills that enable taking safer decisions, thus improving the

management of personal finances. In other words, financial literacy helps individuals with the

process of financial decision-making, as it makes them able to analyze the opportunities

offered and the possible risks that can be generated by guiding them to the choice that will be

the most suitable for their needs, thus easing future financial issues such as debt.

Although research in the area of financial literacy has increased, there is little

consistency in the way this term is defined, as authors approach the topic differently, giving

different connotations to its definition (Hung, Parker, & Yoong, 2009). In addition, studies

highlight the ambiguous use of the term financial literacy, especially in understanding the

differences between the constructs of financial literacy and financial education.

A definition that covers properly this idea, and on which this study is based, is that of

the OECD, which measures financial literacy as a combination of the consciousness,

knowledge, ability, attitude, and behavior necessary to make solid financial decisions, leading

to financial well-being (OECD, 2013). According to OECD (2013), the various terms used to

describe this concept (including in particular financial literacy and financial capability, but

also financial culture and financial insight) could be used relatively interchangeably as they

reflect similar perceptions of the reality they aim to cover. However, they decided to use the

most common international term, “financial literacy”.

The OECD questionnaire is based on a widely accepted working definition of financial

literacy which stresses general behaviors, attitudes and knowledge that could be attained in a

variety of ways. Specifically, the questions cover a mixture of attitudes and knowledge as

well as capturing behavior relating to topics such as money management, planning for short
 6

and longer term financial goals and awareness and choice of financial products (OECD,

2013). With this, the OECD describes financial literacy in terms of three dimensions:

financial knowledge, financial behavior, and financial attitude. As well as, Potrich, Vieira,

Coronel, and Bender Filho (2015) found a valid model for financial literacy measured as a

combination of financial behavior, attitudes and knowledge.

Financial knowledge is a particular kind of human capital that is acquired throughout

life, by learning from issues that affect the ability to effectively manage revenues, expenses,

and savings (Delavande, Rohwedder, & Willis, 2008). Financial behavior is an essential

element of financial literacy and arguably the most important one (OECD, 2013). According

to Atkinson and Messy (2012), the positive results of being financially literate are driven by

behaviors such as expense planning and the construction of financial security. On the other

hand, some behaviors, such as excessive use of credit cards, may reduce financial well-being.

Financial attitudes are established through economic and non-economic beliefs held by a

decision-maker about the outcome of a certain behavior, and therefore are a key factor in the

personal decision-making process (Ajzen, 1991).

Measurement of financial literacy is a complex issue. Lusardi and Mitchell (2011)

defend the argument that, while it is important to assess how people are financially literate, in

practice, it is difficult to explore how people process financial information and make

decisions based on this knowledge. According to Schmeiser and Seligman (2013), the current

commonly used questions in measuring financial literacy have not been rigorously tested to

ensure they are accurately measuring the financial literacy level of the individual. Huston

(2010) conclude that financial literacy has a definition; however, there is no standardized

instrument to measure it.

It therefore seems that there is no "normal" or "usual" way by which researchers or

professionals measure financial literacy. However, this does not mean that there is no
 7

consistency in the way it is measured in the literature and in practice. Annamaria Lusardi and

Olivia Mitchell developed a set of three questions commonly used in various studies. These

questions include three basic concepts: interest rates, inflation, and risk diversification. These

questions are useful here in that they are used in several studies, resulting in a useful

comparison basis between countries.

The lack of standardized measurements and international data as well as requests from

many countries asking for robust measurement of financial literacy at the national level led

the OECD and its International Network for Financial Education to develop a research

instrument to evaluate the financial literacy of people in many countries, based on factors

such as attitudes, behavior, and knowledge associated with global concepts of financial

literacy.

Apart from these instruments, some authors evaluate the dimensions of financial

literacy separately. Van Rooij, Lusardi, and Alessie (2011) examine the relationship between

financial knowledge and retirement planning in Netherlands. Chen and Volpe (1998) and

Johnson (2001) developed evaluative instruments for financial behavior. Another instrument

used in some researches is FL-ABK (financial literacy - attitude, behavior, and knowledge),

elaborated by Shockey (2002), who analyzes the attitudes and behavior of respondents.

Despite differences in the concepts, evaluation, and dimensions inherent to financial

literacy, there is agreement on its importance. According to Schmeiser and Seligman (2013)

and Lusardi and Mitchell (2011), an increase in financial literacy contributes to positive

changes in national prosperity over time.

 8

(b) Gender relationships in financial literacy

It is clear that the differences and inequalities that exist between men and women have

been shaped throughout history by social relations and changes over time. Therefore,

according to a survey from the United Nations Statistics Division (UNSD, 2013), differences

between genders should be discussed.

Specifically in Brazil, this context is aggravated when the study shows a trend observed

throughout the historical series is the growth of the proportion of households headed by

women. In 1995, 23% of the households had women as reference people; while in 2015, this

number reaches 40% (IPEA, 2017). The survey urges reflection on problems and issues

related to women and men in society, consideration of the interest areas in which women and

men do not enjoy the same opportunities and those in which the lives of men and women are

affected in different ways, such as financial literacy.

Gender and financial management relationships have gained increased attention as

understanding an individual’s profile and financial literacy becomes extremely relevant due

to increased financial responsibilities. There is empirical evidence that financial literacy has a

positive impact on individuals’ financial situation and that it is linked to other important

aspects of social welfare. Low financial condition correlates with poor emotional outcomes

for all household members, as well as a lower education level for the children (Shanks &

Danziger, 2010).

Therefore, low financial literacy levels can reduce the ability to accumulate and manage

assets, and ultimately, to ensure a promising financial future (Mottola, 2013). Regarding

gender differences specifically, many authors find that women generally have lower financial

literacy levels than men (Atkinson & Messy, 2012; Mottola, 2013; Agarwalla, Barua, Jacob,

& Varma, 2015; Potrich, Vieira, & Kirch, 2015). In the United States, Lusardi and Mitchell
 9

(2011) find that women are significantly less likely to answer financial questions correctly,

and more likely to say they do not know the answer. On the other hand, women also assess

their own financial literacy level as more conservative, both in developed and developing

countries.

In Brazil, Potrich, Vieira, and Kirch (2015) developed a model that explains

individuals’ financial literacy levels through socioeconomic and demographic variables; they

find that men have higher propensity to be in the group with higher levels of financial

literacy. In addition, Potrich, Vieira, Coronel, and Bender Filho (2015) developed a financial

literacy model adequate for both male and female and male individuals showed a higher level

of financial literacy on average compared to females. Moreover, it seems that the significant

difference between men and women is explained by the fact that men tend to see money as

power and that having money increases their social desirability; women seem to have a more

passive approach toward money (Calamato, 2010). According to Agarwalla, Barua, Jacob,

and Varma (2015), programs for women must strengthen their ability to manage finances and

make them aware of the importance of better planning of consumption.

3. Sample and instrument

This study investigates the resident population of Brazil and by the end of the data

collection period, 2,485 valid surveys were collected, of which 1,062 were from men and

1,423 from women. Note that the sample division between genders was aimed to compare

female behavior to male behavior.

The survey instrument (Appendix 1) uses multiple-choice questions to assess the

respondent’s profile and to measure their financial knowledge. The survey was

complemented by Likert-type 5-point questions assessing the investigated quantitative scales

 10

(for financial attitude and financial behavior). To measure financial literacy level, we used a

multidimensional measure that includes the three constructs suggested by OECD (2013) and

by Potrich, Vieira, Coronel, and Bender Filho (2015): financial attitude, financial behavior,

and financial knowledge.

The first group contains questions that measure financial attitude, adapted from Parrotta

and Johnson (1998). The scale is formed by 15 questions into Likert-type 5-point questions

(1 = strongly disagree to 5 = strongly agree). The questions aim to identify how an individual

self-evaluates financial management. The final eight questions have a reversed Likert scale.

The second group of questions seeks to identify the financial behavior of individuals.

The scale consists of 13 questions and is organized in Likert-type 5-point questions (1 =

never, 5 = always), adapted from Shockey (2002), O’Neill and Xiao (2012) and OECD

(2013). These questions evaluate the financial performance of individuals on two factors:

financial behavior control, with seven questions, and financial behavior of saving, with six

questions.

The last group of questions evaluates the level of financial knowledge, a factor from the

average scores of two groups of multiple-choice questions adapted from Van Rooij, Lusardi,

and Alessie (2011), OECD (2013), Klapper, Lusardi, and Panos (2013), and the National

Financial Capability Study (NFCS 2013), totaling 10 questions. The first set (basic

knowledge) is composed of five questions and aims to measure basic financial abilities, such

as understanding of issues related to inflation, tax rates, and the value of money over time.

The second set (advanced knowledge) is composed of five questions that explore the level of

knowledge in relation to complex financial instruments, such as shares, public bonds, and risk

diversification. The questions’ classification between basic and advanced knowledge is

inspired by Van Rooij, Lusardi, and Alessie (2011), who take into consideration the
 11

questions’ difficulty level. Further, in the stage of instrument validation, the questions were

confirmed by specialist analysis.

Thus, each of the 10 financial knowledge questions was assigned a weighting of 0.5 for

the correct answer, and the financial knowledge index ranges from 0 (score obtained if the

individual mistakes all questions) to 5 (score if the individual correctly answers all

questions). According to this score, respondents were classified as: low level of financial

knowledge (score less than 60% of the maximum), median level of financial knowledge

(between 60% and 79% of the maximum score), and high level of financial knowledge (over

80% of the maximum score). This rating was established by Chen and Volpe (1998). Thus,

financial literacy is measured using the financial attitude (15), financial behavior (13), and

financial knowledge (10) questions.

It is important to note that for the international scales used in this study, adequacy

processes have been developed for the Brazilian context, using the Brislin (1970) model.

Moreover, to validate the scales used, the methodology is based on the Churchill (1979)

model, which is most adequate when searching for problems and solutions concerning

validation of the proposed theoretical model.

4. Data analyses

For the collected data, descriptive statistical analysis and multivariate analysis

techniques were employed, using IBM SPSS Statistics 20.0® and Amos™ software. Initially,

for validation of the constructs, a confirmatory factor analysis (CFA) was used, which,

according to Byrne (2010), is suitable when there is underlying knowledge, whether

theoretical and/or empirical, of a latent variables structure. The validity of the measurement
 12

model was evaluated by checking the convergent validity, reliability, and unidimensionality

of each construct (Hair, Black, Babin, & Anderson, 2010).

The convergent validity of each construct was analyzed according to the magnitude and

meaningful statistics of its standardized coefficients, as well as the absolute fit indices: Chi-

square statistics, root mean square residual (RMR), root mean square error of approximation

(RMSEA), goodness-of-fit index (GFI), and the comparative fit indices: comparative fit

index (CFI), normed fit index (NFI), and Tucker–Lewis index (TLI). There is no consensus

in the literature regarding acceptable values for these indices. For example, for the chi-

square/degrees of freedom, recommendations vary from less than 5 to less than 2. For CFI,

GFI, NFI, and TLI, it is suggested that values be less than 0.95, and for RMR and RMSEA,

the recommendations suggest that values be less than 0.05 and 0.08, respectively (Hooper,

Coughlan, & Mullen, 2008).

A statistical analysis of the values obtained was performed, to verify that the

coefficients of factor loadings were significant at a 95% confidence level. In a CFA, in

addition to observing the coefficients’ statistical significance, it is necessary to observe the

factor loadings magnitude, which must be higher than 0.3 for a sample size of 350 cases or

more (Hair, Black, Babin, & Anderson, 2010). Therefore, variables with factor loadings less

than 0.3 were excluded from the model.

To measure construct reliability, Cronbach's alpha was used; values higher than 0.6

were considered acceptable for the exploratory research. The unidimensionality verification

of the construct was carried out by evaluating the standardized residuals. With absolute

values lower than 2.58, no problems were identified (Hair, Black, Babin, & Anderson, 2010).

Additionally, all models were estimated using a maximum likelihood bootstrap with a sample

size of 1,000, as suggested by Cheung and Lau (2008).

 13

After performing the CFA of the different constructs, financial attitude, behavior, and

knowledge indicators were formed and the cluster analysis applied. This process is known as

conglomerate analysis and is used to classify individuals according to their financial literacy

level. Hierarchical analysis techniques were applied for data analysis. Quadratic Euclidean

distance was used for distance measurement and the Ward method was used as the

agglomeration method. The Ward method was selected because it is one of the most

consistent methods for interval scales. Quadratic Euclidean distance is recommended for the

centroid cluster, and Ward methods were used because they have the advantage of not taking

the square root of the data (Hair, Black, Babin, & Anderson, 2010).

After identifying the cluster to which each individual belongs, the construct’s

descriptive statistics were calculated within each cluster, to find the financial literacy level of

the different groups of the studied population. Further, to check whether there is a significant

difference between the groups, the equality of means test (t-test) and, alternatively, the

Mann–Whitney non-parametric medians test were applied. With these tests, a methodology

for calculation of individual financial literacy level was developed. Finally, frequency

distribution was analyzed (contingency table), to identify the financial literacy level of men

and women, as well as the socioeconomic and demographic profile of the most vulnerable

group.

5. Development of the Financial Literacy Indicator

Following construct validation, we sought to develop the financial literacy indicator to

measure whether individuals have a high or low level of financial literacy. Initially, the

financial attitude and behavior constructs were formed from the weighted average of the

responses to the validated variables of each construct. We used the factor loadings weight.
 14

For instance, the financial attitude construct of a respondent is given by multiplying the factor

loadings weight of the given answer to each question composing the construct. Therefore, the

construct is defined as the weighted average of the responses, with a minimum value of 0 and

a maximum value of 5 points.

Next, to form the financial behavior construct, the weighted average of the factors for

financial control behavior and financial savings behavior was calculated. Again, this was

carried out using the factor loadings weight. For the financial knowledge construct, the basic

financial knowledge and advanced financial knowledge variables were initially calculated.

Then, the financial knowledge indicator was computed from the weighted average using the

factor loadings of the basic and advanced financial knowledge variables.

Finally, to develop the financial literacy indicator, the weighted average of the

construct’s financial attitude, financial behavior, and financial knowledge was obtained.

Thus, in the same way as with the constructs, the financial literacy indicator has a scale of 0

to 5 points.

Next, individuals were classified according to their financial literacy level and a cluster

analysis was applied. According to Malhotra (2011), the obtained clusters must show both an

internal homogeneity (within each group), as well as an external heterogeneity (between

clusters), differing from others groups. For data estimation, the constructs financial attitude,

financial behavior, and financial knowledge were used, and the hierarchical analysis

technique was applied, using the quadratic Euclidean distance measure and the Ward method

of agglomeration.

After identifying the cluster to which the individual belongs, the descriptive statistics of

the constructs within each cluster were calculated, to determine the financial literacy level of

the different groups in the studied sample. Further, to check whether there is a significant
 15

difference between the groups, the average difference test (t-Test) and, alternatively, the

Mann-Whitney non-parametric median difference test were applied.

Finally, an individual financial literacy measure was developed from the calculation of

the quadratic Euclidean distance of the individual from the center of each cluster. From these

distances, it is possible to determine which cluster comes closer to the individual and

consequently it becomes possible to classify whether the individual has a low or high level of

financial literacy.

6. Results and discussions

(a) Sample profile

By the end of the survey, 2,485 valid samples were collected, meeting the previously

established minimum sample required for all municipalities surveyed. Of this total, 1,062

responders were men and 1,423 women, representing, respectively, 42.74% and 57.26% of

the final sample.

Regarding age and marital status, 26.70% of women and 27.12% of men were under the

age of 23, and 46.04% of women were married and 48.67% of men were single. Most

respondents stated they were not studying (64.30% of women and 65.79% of men) and they

have no dependents (52.38% of women and 56.77% of men).

Another question referred to the participant's education level, considering that this can

affect financial education level. For this question, 42.49% of women and 46.41% of men

stated that they had completed high school and 26.66% of women and 25.45% of men

responded that they had finished university. Most women said that their parents had

completed only primary education (58.16%), as did 47.43% of men. Finally, 39.77% of
 16

women are salaried employees and 9.47% are not working. On the other hand, 36.15% of

men are salaried employees and 11.66% are self-employed.

Concerning aspects such as income, 33.64% of women and 30.43% of men reported

receiving between R$868.01 and R$1,738.00 per month. On the other hand, 10.95% of

women and 8.13% of men do not have their own income. Only 16.64% of women earn more

than R$2,604.00, compared to 27.37% of men, demonstrating the disparity that still exists in

the labor market when it comes to gender. As to average monthly family income, 19.96% of

women and 19.90% of men reported belonging to a family with an income between

R$1,738.01 and R$2,604.00. Some 23.92% of women reported having a family income

between R$868.01 and R$1,738.00, and 17.99% of men reported family income between

R$2,604.01 and R$3,472.00.

(b) Validation of the constructs and integrated model

The first objective of this study was to validate the constructs used as proxies to

measure the financial literacy of individuals through the CFA. Note that the financial

knowledge construct did not pass through the validation process, as it was constructed from

nominal variables (right or wrong answer).

To validate the constructs used as proxies to measure the financial literacy of

individuals through the CFA, we examine the relationships between the observed variables

and their constructs, performing the estimation using the maximum likelihood method.

Therefore, following the strategies initially proposed, the measurement model of the

constructs was set (Table 1).

 17

Table 1

Fit indices of model for financial behavior and financial attitude constructs

Financial Behavior Financial Attitude

Fit Indices
Proposed Final Proposed Final
Chi-square (value) 2758.784 34.609 1109.780 129.549
Chi-square (p-value) <0.001 0.054 <0.001 0.055
Degrees of freedom 90 10 64 28
Chi-square / Degrees of freedom 30,653 3,461 17,340 4,627
GFI – Goodness-of-fit index 0.821 0.996 0.932 0.991
CFI – Comparative fit index 0.738 0.996 0.928 0.992
NFI – Normed fit index 0.732 0.995 0.924 0.990
TLI – Tucker–Lewis index 0.695 0.992 0.912 0.985
RMR – Root mean square residual 0.077 0.009 0.064 0.029
RMSEA – Root mean square error of
0.109 0.031 0.081 0.038
approximation
Cronbach’s alpha 0.814 0.850 0.894 0.897

Initial results indicated that both models are inadequate. Therefore, a strategy of

excluding variables with no significant standardized factor loadings (values below 0.3) was

adopted. Note that the exclusion was individual, that is, one variable at a time, in ascending

order (from the smallest to the largest coefficient). For each exclusion, a new model was

estimated. In addition, correlations between the error variables were inserted, which was

suggested by the software and which made theoretical sense.

Following these changes, the two models now provided adequate adjustment: i)

convergent validity, since the GFI, CFI, NFI, and TLI indexes were above 0.95 and RMR and

RMSEA indices were less than 0.05; ii) reliability was higher than 0.6; and iii)

unidimensionality, since all the standardized residual values were below 2.58. In addition, the

chi-square at the level of 6% was no longer significant, confirming the estimated and

observed matrix adjustments.

After observing each of the constructs individually along with their training variables,

we sought to build the integrated model proposed for this study, which combined the

validated measurement models and the structural model. The integrated model was evaluated
 18

using the adjustment rates and the statistical significance of the estimated regression

coefficients. Therefore, to confirm whether the initial model was suitable, the adjustment

indices were analyzed and are shown in Table 2.

Table 2
Fit indices of model for financial literacy
Model – Financial Literacy
Fit Indices
Chi-square (value) 562.435 TLI – Tucker–Lewis Index 0.975
Chi-square (p-value) <0.001 RMR – Root mean square residual 0.040
RMSEA – Root mean square error of
Degrees of freedom 147 0.034
approximation
Chi-square / Degrees of freedom 3.826 AIC – Akaike's information criterion 688.435
GFI – Goodness-of-fit index 0.977 BIC – Bayes information criterion 1055.022
CFI – Comparative fit index 0.980 ECVI – Expected cross-validation index 0.277
NFI – Normed fit index 0.974

In Table 2, it appears that the initially proposed model for measuring financial literacy

level had adequate adjustment indices, as the CFI, GFI, NFI, and TLI indexes were above

0.95 and RMR and RMSEA indexes were below 0.05. However, the chi-square value

remained significant, but the chi-square / degrees of freedom ratio was below the

recommended range (below 5). This significance is justified, since one of the problems

shown by the chi-square statistic is its sensitivity to sample size; as the sample size increases,

the pattern tends to show significant results (Byrne, 2010; Kline, 2011).

Thus, it appeared that the proposed initial model had adequate adjustment indices; it

was not necessary to make modifications. Figure 1 shows the final integrated model with the

coefficients standardized and significance of the relationship.

 19

Figure 1
Final model for financial literacy

Notes: * p < 0.01;

¹ z value not calculated where the parameter is set at 1, due to model requirements. For simplicity,
correlations among the errors are not shown.

All the proposed relationships checked were significant. Specifically, note that, in the

final measurement model of the financial attitude construct, 7 of the 15 variables initially

established were kept, requiring inclusion of four additional correlations among the variable

errors, intended for adjustment of the model. Analysis of the final model composition shows

that the financial attitude construct can be measured mainly by variables, which address

whether individuals consider it absolutely important to have a written budget for successful

personal financial management and whether it is essential to plan for a possible loss of

earnings from a family member, respectively.

The final measurement model of the financial behavior construct was formed by 11 of

the 13 variables originally proposed, from which 5 measure financial control behavior and 6

measure financial savings behavior. Additionally, the variables have the most impact in

measuring the financial behavior of individuals’ control level, highlighting how frequently
 20

each behavior occurs with respect to budget planning and analysis of the budget before

making a large purchase.

As for financial savings behavior, note that all variables showed high standardized

coefficients, how often the individual maintains a money reserve from monthly income for

future needs. Moreover, note that both dimensions strongly impact the financial behavior

construct. However, the magnitude of financial control behavior shows a greater impact

compared to financial savings behavior. Those who frequently control their finances have a

higher tendency toward more adequate financial behavior.

Finally, observe that the financial behavior construct has the most impact in the

integrated model of financial literacy, followed by the financial attitude construct, and finally,

the financial knowledge construct. This result corroborates Klapper, Lusardi, and Panos

(2013) in that financial behavior is a genuine and determining element of financial literacy.

(c) Development of the Financial Literacy Indicator

From the validated models, and to achieve the objective of developing a financial

literacy indicator, the financial attitude, financial behavior, and financial knowledge

indicators were initially built to form the financial literacy indicator with four steps.

The Step 1, with the answers of respondents to the questions marked in bold in the

financial literacy questions (Appendix 1), the variables are coded as follows:

 FINANCIAL ATTITUDE – Q01, Q02, Q03, Q04, Q05, Q06 and Q07 (Likert

scale): Strongly disagree = value 1; Disagree = value 2; Indifferent = value 3;

Agree = value 4; and Completely agree = value 5.

 FINANCIAL BEHAVIOR – Q16, Q18, Q19, Q20, Q21, Q22, Q23, Q25, Q26,

Q27 and Q28 (Likert scale): Never = value 1; Almost never = value 2;

Sometimes = value 3; Almost always = value 4; and Always = value 5.

 21

 FINANCIAL KNOWLEDGE – Q29 to Q38 (right or wrong questions): for the

right questions = value 1; and for the wrong questions = value 0.

After coding the variables, the Step 2 calculates of the standardized measures of

financial literacy. For each construct, the factor loading weights were obtained, then the

weighted average of these weights was calculated. Thus, construct financial attitude was

calculated for each respondent, from the weighted sum: ATTIT = [0.125*Q01 + 0.148*Q02 +

0.157*Q03 + 0.154*Q04 + 0.149*Q05 + 0.150*Q06 + 0.117*Q07]. Construct financial

behavior, comprising financial behavior control and savings, was calculated from weighted

sum [0.541*CONTROL + 0.459* SAVINGS], with the variable CONTROL obtained via the

weighted sum: CONTROL = [0.177*Q16 + 0.242*Q19 + 0.197* Q20 + 0.173*Q21 +

0.211*Q23] and the variable SAVINGS via the weighted sum: SAVINGS = [0.176*Q18 +

0.173*Q22 + 0.172* Q25 + 0.155*Q26 + 0.156*Q27 + 0.169*Q28].

In addition to financial attitude and financial behavior, the index of the financial

knowledge construct was also calculated, comprising the three proxies measuring financial

literacy. As previously mentioned, it is known that the financial knowledge construct consists

of 10 multiple-choice questions, 5 on basic financial knowledge and 5 on advanced

knowledge. For each of the questions, a value of 0.5 was given for the correct answer and 0

for incorrect answers. Thus, the basic financial knowledge construct was obtained by the sum

[Q29 + Q30 + Q35 + Q36 + Q37] and the advanced financial knowledge construct by [Q31 +

Q32 + Q33 + Q34 + Q38]. Further, for analysis purposes, these constructs were standardized

according to the weights they represent in the financial knowledge construct, using the

expression [0.506*BASIC+ 0.494*ADVANCED].

After finding the variables corresponding to the attitude, behavior, and financial

knowledge constructs, we performed Step 3 to develop weighted measures of financial

literacy. For this, the suns of attitude, behavior, and financial knowledge constructs (Step 2)
 22

were then multiplied by the value 0.337; 0.424 and 0.238, respectively, to determine the

weight that each construct has on financial literacy.

Having considered the three constructs, cluster analysis was applied to classify

individuals’ financial literacy level as either low or high. For this purpose, hierarchical cluster

analysis was performed using the Ward method, resulting in two clusters. The first one is

referred to as Cluster 0, representing individuals who have a low financial literacy level

(60.8%), and the second one is called Cluster 1, representing those with a high level of

financial literacy (39.2%). Table 3 shows the descriptive statistics of the weighted constructs

according to the cluster distribution.

Table 3

Descriptive statistics of the constructs according to cluster distribution

Cluster 0 Cluster 1
Test
n=1285 (60.8%) n=830 (39.2%)
Test t Mann–
Financial Literacy Financial Literacy
Constructs Whitney
Low Level High Level
Standard Standard
Mean Median Mean Median Sig. Sig.
Deviation Deviation
Standardized

Financial behavior 3.107 3.099 0.843 4.140 4.127 0.479 <0.001 <0.001
scale

Financial knowledge 2.084 2.012 1.338 3.872 4.000 0.612 <0.001 <0.001
Financial attitude 4.107 4.000 0.598 4.326 4.277 0.475 <0.001 <0.001

Financial behavior 1.344 1.314 0.357 1.756 1.750 0.203 <0.001 <0.001
Weighted
scale

Financial knowledge 0.496 0.479 0.318 0.922 0.952 0.146 <0.001 <0.001

Financial attitude 1.384 1.348 0.201 1.458 1.441 0.160 <0.001 <0.001

Analysis of the descriptive statistics, we confirm that individuals with a high level of

financial literacy (Cluster 1) showed an adequate level of financial behavior. On the

standardized scale (1-5 points), the responses are equivalent to the “almost always” option

(point 4 on the scale). Concerning respondents with a low level of financial literacy (Cluster

0), we have an average result equivalent to the option “sometimes” on the scale (point 3).

Therefore, we obtained a significant difference between the groups, with high frequency of
 23

good financial behavior in individuals with high levels of financial literacy, and median

frequency in the low level of financial literacy group. The main positive points found regard

financial control behavior, such as paying bills on time and analyzing the budget before

making a large purchase. Those who had a lower frequency of financial savings behavior

referred to issues such as having a financial reserve equal to or greater than three times

monthly expenses, and managing to save money in the past 12 months.

In the financial knowledge evaluation, the performance achieved by the respondents

resulted in a standardized average of 2,084 for individuals with a low level of financial

literacy and 3,872 for those with a high level. Accordingly, on average, respondents with low

levels of financial literacy answered correctly 41.69% of the proposed questions, and those

with a high level achieved an average rate of 77.44% correct answers. Based on the

classification proposed by Chen and Volpe (1998), individuals classified in Cluster 1

obtained a median level of financial knowledge of between 60% and 79%, which is, however,

very close to the high level (above 80%). Those belonging to Cluster 0 showed a low level

(below 60%). This result is concerning, as most of the surveyed sample (60.8%) shows a low

level of financial literacy, and an understanding of interest rates, inflation, and the value of

money over time is essential to performing daily financial transactions.

However, the low level of financial literacy found is not exclusive to this research. In

recent years, multiple researchers have investigated financial knowledge levels, both among

students and the general population. These researchers obtained quite worrying results, with

unsatisfactory levels of knowledge, both concerning personal financial management issues

and more specific questions, such as credit, loans, savings, and investments (Lusardi,

Mitchell, & Curto, 2010; Van Rooij, Lusardi, & Alessie, 2011).

Sekita (2011), in Japan, revealed that 62% of surveyed students said they did not have

enough knowledge to answer questions about interest rates, the combined effect of interest
 24

rates and inflation, and investment risk diversification. Lusardi, Mitchell, and Curto (2010),

in a study on US residents, also found low levels of financial literacy among young people,

with only 27% answering correctly questions regarding these three issues.

Financial attitude was not a decisive factor for the financial literacy classification, as

the other two constructs were preponderant. Since both groups showed a good level of

financial attitude, their responses in general associate with the “agree” option (point 4 on the

scale). Afterwards, to develop an indicator defining whether an individual has a high or low

level of financial literacy, that is, determining the financial literacy indicator, we used the

squared Euclidean distance in relation to the cluster center, while having the three constructs

representing the proxies as variables, to measure financial literacy (financial attitude,

financial behavior, and financial knowledge). Equations 1 and 2 show how to estimate

individuals’ responses with a distance to the Cluster 0 center (Equation 1) and the Cluster 1

center (Equation 2). From these equations, it can be determined whether the respondent is

closest to Cluster 0 (D0) or Cluster 1 (D1).

(1)

(2)

Finally, after inserting the results of Step 3 in equations 1 and 2, we find the level of

financial literacy of the individuals (Step 4). Therefore, if the individual shows > , the

individual is considered to have a high financial literacy level. But if the response obtained is

< , the individual is considered to have a low financial literacy level, because the

distance of his responses to the Cluster 0 center is smaller than the distance from the Cluster

1 center. To demonstrate the results in a more applied way, a methodology, step by step, for

calculating the financial literacy index was developed (Chart 1).

 25

Chart 1
Methodology for calculating Financial Literacy Indicator

METHODOLOGY FOR CALCULATING FINANCIAL LITERACY INDICATOR

Step 1: With the answers of respondents to the questions marked in bold in the financial literacy
questions (Appendix 1), the variables are coded:
FINANCIAL ATTITUDE – Q01, Q02, Q03, Q04, Q05, Q06 and Q07 (Likert scale):
Strongly disagree = value 1
Disagree = value 2
Indifferent = value 3
Agree = value 4
Completely agree = value 5
FINANCIAL BEHAVIOR – Q16, Q18, Q19, Q20, Q21, Q22, Q23, Q25, Q26, Q27 and Q28 (Likert scale):
Never = value 1
Almost never = value 2
Sometimes = value 3
Almost always = value 4
Always = value 5
FINANCIAL KNOWLEDGE – Q29 to Q38 (right or wrong questions):
For the right questions = value 1
For the wrong questions = value 0

Step 2: Development of standardized measures of financial literacy:

FINANCIAL ATTITUDE
ATTIT = [0.125*Q01 + 0.148*Q02 + 0.157*Q03 + 0.154*Q04 + 0.149*Q05 + 0.150*Q06 + 0.117*Q07]

FINANCIAL BEHAVIOR
Control Financial Behavior
CONTROL = [0.177*Q16 + 0.242*Q19 + 0.197* Q20 + 0.173*Q21 + 0.211*Q23]
Savings Financial Behavior
SAVINGS = [0.176*Q18 + 0.173*Q22 + 0.172* Q25 + 0.155*Q26 + 0.156*Q27 + 0.169*Q28]
BEHAV = [0.541*CONTROL + 0.459* SAVINGS]

FINANCIAL KNOWLEDGE
Basic Financial Knowledge
BASIC = [Q29 + Q30 + Q35 + Q36 + Q37]
Advanced Financial Knowledge
ADVANCED = [Q31 + Q32 + Q33 + Q34 + Q38]
KNOW = [0.506*BASIC + 0.494*ADVANCED]
Step 3: Development of weighted measures of financial literacy:
wATTIT = 0.337*ATTIT
wBEHAV = 0.424*BEHAV
wKNOW = 0.238*KNOW

Step 4: Enter the results in the formulas:

2 2 2
0 = 1.384 + 1.344 + 0.496
2 2 2
1 = 1.458 + 1.756 + 0.922
Step 5: Analysis and decision criteria:
- If 0 > 1  The individual is considered to have HIGH level of financial literacy
- If 0 < 1  The individual is considered to have LOW level of financial literacy
 26

It is also noteworthy that the proposed model measuring individuals’ financial literacy

is supported by several authors, establishing that financial literacy cannot be determined

using single and isolated measures; one should rather include a more complex analysis of

different constructs together (Moore, 2003; Atkinson & Messy, 2012).

According to Moore (2003), when an individual demonstrates better financial

knowledge and financial attitude as well as more positive financial behavior, it is expected

that the individual would be more financially literate and consequently, more effective in

financial management. Furthermore, according to Agarwalla, Barua, Jacob, and Varma

(2015), insights on these dimensions may separately provide more valuable information than

those provided by a single construct determining financial literacy only.

(d) Financial literacy level: gender x socioeconomic and demographic variables

For an analysis of the socioeconomic and demographic variables and financial literacy,

Table 4 shows the frequency distributions of the financial literacy variable together with the

gender variable. The last column of this table refers to Pearson chi-square association

(p-value in brackets) between the following pair: financial literacy x gender.

Table 4

Contingency table: financial literacy x gender

Financial Literacy Low (0) High (1)

Chi-square
Variable % (line) % (line) Pearson [p-value]
Female 66.3 33.7 35.749
Gender
Male 53.5 46.5 [<0.001]

The results indicate a statistically significant association at the 1% level between

financial literacy and gender. There is a greater proportion of individuals with a high level of
 27

financial literacy among men (46.5%) than among women (33.7%), confirming a priori

expectations and previous studies.

Among the reasons found for the lower levels of literacy among women, the long

struggle in the search for a larger social space, autonomy, and recognition are highlighted. In

this regard, contrary to a common view, women have always worked in housekeeping,

contributing to their family's survival and performing tasks that make them part of the unpaid

workforce (Bruschini & Rosemberg, 1982).

Nowadays, women are not restricted to household activities only; they work in virtually

all social spheres. There is, however, still prejudice. To alleviate this condition, women are

increasingly seeking to overcome various pejorative representations about the female gender.

Women are gaining ground in the labor market, beginning to participate increasingly in the

household’s financial management, and often becoming head of household. Between 2000

and 2010, according to the IBGE census, there was an increase in families with women as

head of household (from 22.2% to 37.3%), and a decrease in the percentage of responsible

men by earning family with the spouse’s presence, from 95.3% to 92.2%. The reasons for this

can be attributed to a change in values about women’s role in society, and factors such as

their entry into the labor market, as well as increasing education combined with reduced

fertility (IBGE, 2014).

Based on these factors, women’s relationship with financial management has gained in

importance. Therefore, having knowledge of these women’s profiles and their financial

literacy levels becomes extremely relevant, due to the increase in their financial

responsibilities. This is mainly because there is empirical evidence that financial literacy

positively impacts individuals’ financial attitudes, as well as the skills to adequately keep a

budget, save money and control spending (Moore, 2003), and knowledge of how to control

debts (Campbell, 2006), and accumulate wealth (Stango & Zinman, 2009). Therefore, low
 28

levels of financial literacy among women may be hindering their ability to accumulate and

manage assets, and ultimately ensure a promising financial future (Mottola, 2013).

Finally, to identify socioeconomic and demographic variables that differentiate the

most vulnerable group, in this case women with low and high levels of financial literacy, a

contingency table analyzing each profile according to financial literacy level was developed

(Table 5).

Table 5

Female profile according to financial literacy level

Financial Literacy Low (0) High (1)

Chi-square
Variable % (line) % (line) Pearson [p-value]
Single 58.0 42.0 9.583
Marital status
Married 47.8 52.2 [0.002]
Primary school 74.5 25.5
High school 75.1 24.9
Technical course 67.7 32.3 59.125
Education Graduate level 58.5 41.5 [<0.001]
Specialization or MBA 46.0 54.0
Master/Doctorate/Post-doctorate 38.2 61.8
No income 69.6 30.4
Up to R$868.00 74.2 25.8
Between R$868.01 and R$1738.00 69.0 31.0
Between R$1738.01 and R$2640.00 64.7 35.3 53.212
Personal income Between R$2640.01 and R$3472.00 56.6 43.4 [<0.001]
Between R$3472.01 and R$5208.00 44.2 55.8
Between R$5208.01 and R$7812.00 33.3 66.7
Between R$7812.01 and R$10616.00 12.5 87.5
More than R$10616.00 36.4 63.6
Up to R$868.00 79.3 20.7
Between R$868.01 and R$1738.00 76.6 23.4
Between R$1738.01 and R$2640.00 71.8 28.2 62.965
Between R$2640.01 and R$3472.00 60.8 39.2 [<0.001]
Household income
Between R$3472.01 and R$5208.00 63.0 37.0
Between R$5208.01 and R$7812.00 50.0 50.0
Between R$7812.01 and R$10616.00 35.2 64.8
More than R$10616.00 44.8 55.2
Total 66.3 33.7

From the presented measures, observe that there is a statistically significant association

at the 1% level between women’s financial literacy and the variables for marital status,

education, personal income, and household income. Among single women, there is a smaller
 29

share with a high financial literacy level (42.0%), than among married women (52.2%). This

is a result similar to previous studies and signals possible reverse causality between these

variables. According to Brown and Graf (2013), single people have a significant tendency

toward lower levels of financial literacy, compared to married individuals. In general, when

people have a low financial literacy level, they run the risk of making bad financial decisions

that, in the long term, may result in debts, endangering their relationships (Calamato, 2010).

Confirming this evidence, Dew (2008) finds that consumer debt is a major threat to marital

satisfaction, and for this reason married individuals tend to have higher levels of financial

literacy.

With respect to education and income levels, as expected, the proportion of women

with high levels of financial literacy grows along with education levels, personal income, and

household income. Note that women without a university education show low levels of

literacy, in contrast to those who have completed graduate levels, who show a high level of

financial literacy. In this sense, there is an increase in the proportion of women with a high

financial literacy level, when passing from technical training to university education (9.2%

increase), from university to a specialization or MBA (12.5% increase), and from this latter

level to a master/doctoral/post-doctoral level (7.8% increase), indicating a strong relationship

between education and financial literacy. Note that having a university degree alone is not

enough for women to show a high level of financial literacy. This fact is quite concerning,

since, a priori, a more appropriate level of financial literacy for those with a college

education is expected.

Finally, there is an increase in the proportion of high financial literacy for women with

increased levels of own and family incomes. Women with income levels up to R$3,472.00

and family income up to R$5,208.00 show a low level of financial literacy. From this income
 30

level on, they begin to show a higher level of financial literacy. This finding again indicates a

strong relationship between income levels and financial literacy.

In this context, these results become important because of the many personal financial

responsibilities that individuals take over a lifetime. As they increasingly manage family

budgets subject to income restrictions, purchase products and services, monitor financial

accounts, deal with credit cards, and save or invest for a future event such as a child's college

or retirement. However, the difficulty in knowing everything a person should know about

personal finances in a constantly changing and increasingly complex financial world is a

huge challenge, even for the most educated adults (Allgood & Walstad, 2016).

Corroborating, Atkinson and Messy (2015) emphasize that the skills of being

financially literate not only facilitate the financial inclusion of individuals, but it is also vital

for effective money management and long-term financial planning, and such behaviors can

improve financial well-being. Furthermore, financial literacy is associated with different

behavioral factors, for example, the individual who is financially literate decreases his

compulsive buying and materialistic behavior, in addition to his propensity for indebtedness

(Cakarnis & D'Alessandro, 2015). In other words, the importance of financial literacy is

greater than helping the finances of individuals, and it contributes to other behavioral

problems that affect the financial health of society.

Specifically, the differences in financial literacy among different socioeconomic and

demographic groups, such as gender, may be important to focus on the most vulnerable and,

for example, to create national strategies targeting these groups. In addition, the availability

of data on financial literacy can also better explain the causes and consequences of financial

illiteracy and this information can help enrich the framework for studying financial decision

making (Lusardi, 2015).

 31

This research becomes even more relevant after the publication presented by the

OECD, in October 2016, when they released the report “OECD-INFE International Survey of

Adult Financial Literacy Competencies”, which it presents the results of a study on the

knowledge, attitude and financial behavior of 51,650 adults from 30 countries, including

Brazil for the first time. In this sense, although the results indicated that the level of financial

literacy is low worldwide, Brazil's score was even lower, standing at 1.2 percentage points

below the world average (OECD, 2016). This report highlights the urgent need for more

effective action to minimize the effects that low levels of financial literacy can have on

individuals' lives and, consequently, on the economy.

7. Conclusions

This study aims to develop, in the Brazilian context, an indicator assessing financial

literacy level while analyzing gender differences, as well as identifying the socioeconomic

and demographic profile of the most vulnerable group. The study innovates by shaping

financial literacy as a multidimensional construct, and by proposing an effective methodology

to measure the level of the individuals’ financial literacy.

In a preliminary analysis, the majority of respondents, both men and women, showed a

low level of financial literacy. However, from the bivariate association measures, the

existence of a significant relationship between financial literacy and gender can be observed;

the proportion of male individuals is higher among those with a high level of financial

literacy, confirming a priori expectations and previous studies.

Further, among the most vulnerable group, in this case female individuals, a significant

association between financial literacy and marital status, education, personal income, and

household income variables was found. These results are summarized as follows: Single
 32

women with lower levels of education and low family and own incomes have the highest

tendency toward membership in the group with low financial literacy levels.

These findings confirm the urgent need to develop effective policies to minimize the

problem of financial illiteracy. One of the possible actions to be taken pertains to inclusion of

financial management disciplines and market finance concepts in all undergraduate courses,

regardless of the teaching discipline, since even women with university degrees show low

levels of financial literacy. Another possible measure concerns the development and adoption

of educational programs to promote personal financial literacy in all sectors of society. Some

actions in this direction are being promoted, specifically by the central bank and federal

government of Brazil, through the National Strategy of Financial Education. However, we

suggest more specific methodologies to promote, for example, extension projects aiming to

conduct financial literacy courses, not just focused in teaching financial concepts, but with

tips and practices for the improvement of attitudes and financial behaviors. A strategy

construction aimed at adopting disciplines and contents focused on financial literacy in the

early levels of education could, in the long run, make children more prepared for financial

management and reduce gender inequalities, before individuals become adults and before

they become responsible for management of their own resources.

In addition, from the proposed indicator, it is possible to know the individual financial

literacy level, enabling financial institutions to propose strategies and financial products more

suited to their customers’ profiles. From the government’s point of view, such an indicator

could increase awareness about the most vulnerable groups, allowing government to focus its

actions on promoting improvements in financial literacy. The results suggest that greater

efforts should be made to support single women who have lower levels of education and

income. Potrich, Vieira, and Kirch (2015) show that such profile variables are determinants

of the level of financial literacy. A next step for future research would be to assess the
 33

moderating effect of gender on the relationship between financial literacy and these profile

variables, since one can argue that the effects of education and income on financial literacy,

for example, vary with gender.

As this research’s main contribution, we highlight that this study is a pioneer in the

Brazilian context by proposing an indicator identifying whether individuals have a low or a

high level of financial literacy. With this knowledge, one can more easily understand the

individual’s profile and develop actions to improve financial literacy, prevent excessive debt,

and to help individuals achieve financial well-being, focusing specifically on the profile that

shows major deficiencies: single women with low levels of education and income.

The results are important for the development of public policies and other agents

interested in the subject, since financial literacy goes beyond the fact of impacting the

financial health of those who have it. Financial institutions will also have the opportunity to

take advantage of this information in order to better understand the financial literacy of their

clients and they would build products more suited to each socioeconomic and demographic

profile in order to tailor their products, services or operations to the profile of each client

(“Process of Suitability”).

Therefore, one of the main results of this research refers to the awareness of the public

and private financial agents that the analysis of the behavioral profile of individuals should be

incorporated into the existing models for an improvement in the prediction of default of

families. It is also worth noting that financial uncontrolled situation brings consequences such

as social marginalization and exclusion, as well as physical and mental illnesses, which in the

future may not only aggravate the level of defaults in the country, but also increase

expenditures for psychological and health care public.

These conclusions reaffirm the urgency and need for effective action to be taken to

minimize the problem of financial illiteracy and gender inequality. Brazil has adopted public
 34

policies to reduce gender inequality and women's empowerment, such as, for example, the

money from the Bolsa Família Program being paid primarily to women. However, if women

are less financially literate, resource management could be less efficient, further ratifying the

importance of adopting public policies to reduce inequality in financial literacy between men

and women.

In this sense, some actions have already been undertaken, especially by the Central

Bank and the Federal Government in Brazil, such as the proposal to include the theme of

financial education in the National Curricular Common Base, which may reduce gender

inequalities caused by social reasons. Since formal education should treat everyone equally

and over the long term, gender inequalities tend to decline.

These results are ratified by the OECD, which suggests that one of the measures to

minimize the effects of low levels of financial literacy is that schools can also help children

develop the skills and attitudes that will help them in this process by encouraging habits and

behaviors positive, how to make a spending plan, save and plan ahead (OECD, 2016). In

summary, this paper brings conceptions and developments that lead beyond the core theme of

financial literacy. With the results it becomes possible to think about the subject of financial

citizenship, a concept that refers to rights and duties related to the financial life of the citizen,

to which financial education has one of its pillars, besides the protection of the consumer of

services financial and the financial inclusion of the population. In this sense, in order to

achieve financial citizenship, a country must, in addition to guaranteeing consumer protection

of financial services and financial inclusion, promote quality financial education.

 35

References

Agarwalla, S. K., Barua, S., Jacob, J., & Varma, J. R. (2015). Financial Literacy among

Working Young in Urban India. World Development, 67, 101-109.

Ajzen, I. (1991). The theory of planned behavior. Organization Behavior Human Decision

Process, 50, 179-211.

Allgood, S., & Walstad, W. B. (2016). The Effects of Perceived and Actual Financial

Literacy on Financial Behaviors. Economic Inquiry, 54(01), 675-697.

Atkinson, A., & Messy, F. (2012). Measuring financial literacy: results of the OECD /

International Network on Financial Education (INFE) pilot study.

http://dx.doi.org/10.1787/5k9csfs90fr4-en

Brislin, R. W. (1970). Back-Translation for Cross-Cultural Research. Journal of Cross-

Cultural Psychology, 1(03), 1-185.

Brown, M., & Graf, R. (2013). Financial literacy and retirement planning in Switzerland.

Numeracy, 6(02), 6.

Bruschini, C., & Rosemberg, F. (1982). Trabalhadoras do Brasil. São Paulo: Brasiliense.

Bucher-Koenen, T., Lusardi, A., Alessie, R., & Van Rooij, M. (2017). How Financially

Literate Are Women? An Overview and New Insights. Journal of Consumer Affairs,

51(02), 255-283.

Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications,

and programming. 2 Ed. New York: Routledge.

 36

Cakarnis, J., & D’Alessandro, S. P. (2015). Does knowing overcome wanting? The impact of

consumer knowledge and materialism upon credit card selection with young

consumers. Young Consumers, 16(01), 50-70.

Calamato, M. P. (2010). Learning Financial Literacy in the Family. Unpublished master’s

thesis. The Faculty of the Department of Sociology, San José State University.

Campbell, J. (2006). Household finance. Journal of Finance, 61(04), 1553-1604.

Chen, H., & Volpe, R. P. (1998). An analysis of personal financial literacy among college

students. Financial Services Review, 7(02), 107-128.

Cheung, G. W., & Lau, R. S. (2008). Testing mediation and suppression effects of latent

variables: Bootstrapping with structural equation models. Organizational Research

Methods, 11(02), 296-325.

Churchill, G. A. A. (1979). Paradigm for Developing Better Measures of Marketing

Constructs. Journal of Marketing Research, 16(01), 64-73.

Delavande, A., Rohwedder, S., & Willis, R. J. (2008). Retirement Planning and the Role of

Financial Literacy and Cognition. Michigan Retirement Research Center, Working

Paper 2008-190.

Dew, J. (2008). Debt change and marital satisfaction change in recently married couples.

Family Relations, 57(01), 60-71.

Garver, N. S., & Mentzer, J. T. (1999). Logistics Research Methods: Employing Structural

Equation Modeling to Test for Construct Validity. Journal of Business Logistics,

20(01), 33-57.
 37

Gerardi, K., Goette, L., & Meier, S. (2010). Financial Literacy and Subprime Mortgage

Delinquency: Evidence from a Survey Matched to Administrative Data, Federal

Reserve Bank of Atlanta, Working Paper Series No. 2010-10.

Hair, J. R., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate Data

Analyses. 7 Ed. New Jersey: Pearson.

Hooper, D., Coughlan, J., & Mullen, M. R. (2008). Structural Equation Modelling:

Guidelines for Determining Model Fit. The Electronic Journal of Business Research

Methods, 6(01), 53-60.

Hung, A. A., Parker, A. M., & Yoong, J. (2009). Defining and measuring financial literacy.

Social Science Research Network.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1498674.

Huston, S. J. (2010). Measuring Financial Literacy. Journal of Consumer Affairs, 44(02),

296-316.

Instituto Brasileiro de Geografia e Estatística (IBGE). (2014). Estatísticas de gênero: uma

análise dos resultados do censo demográfico 2010.

http://biblioteca.ibge.gov.br/visualizacao/livros/liv88941.pdf

Instituto de Pesquisa Econômica Aplicada (IPEA). (2017). Retrato das Desigualdades de

Gênero e Raça – 1995 a 2015.

http://www.ipea.gov.br/portal/images/stories/PDFs/170306_retrato_das_desigualdades_

de_genero_raca.pdf

Johnson, A. C. (2001). Evaluating a financial assessment tool: the financial checkup.

Dissertation, Utah University, United States.

 38

Klapper, L., Lusardi, A., & Panos, G. A. (2013). Financial literacy and its consequences:

Evidence from Russia during the financial crisis. Journal of Banking & Finance,

37(10), 3904-3923.

Kline, R. B. (2011). Principles and practice of structural equation modeling. 3 Ed. New

York: The Guilford Press.

Lusardi, A. (2015). Financial literacy: Do people know the ABCs of finance? Public

Understanding of Science, 24(03), 260-271.

Lusardi, A., & Mitchell, O. S. (2011). Financial literacy and retirement planning in the

United States. Journal of Pension Economics and Finance, 10(04), 509-525.

Lusardi, A., Mitchell, O. S., & Curto, V. (2010). Financial literacy among the young. Journal

of Consumer Affairs, 44(02), 358-380.

Malhotra, N. K. (2011). Pesquisa de marketing: uma orientação aplicada. 6 Ed. Porto

Alegre: Bookman.

Moore, D. (2003). Survey of Financial Literacy in Washington State: Knowledge, Behavior,

Attitudes, and Experiences. Social and Economic Sciences Research Center,

Washington State University.

Mottola, G. R. (2013). In our best interest: Women, financial literacy, and credit card

behavior. Numeracy, 6(02), 4.

National Financial Capability Study (NFCS). (2013). Financial Capability in the United

States - Report of Findings from the 2012 National Financial Capability Study.

Financial Industry Regulatory Authority (FINRA) Investor Education Foundation.

 39

http://www.usfinancialcapability.org/downloads/NFCS_2012_Report_Natl_Findings.p

df

O’Neill, B., & Xiao, J. (2012). Financial Behaviors Before and After the Financial Crisis:

Evidence from an Online Survey. Journal of Financial Counseling and Planning,

23(01), 33-46.

Organisation for Economic Co-Operation and Development (OECD). (2013). Financial

literacy and inclusion: Results of OECD/INFE survey across countries and by gender.

OECD Publishing.

http://www.oecd.org/daf/fin/financialeducation/TrustFund2013_OECD_INFE_Fin_Lit

_and_In cl_SurveyResults_by_Country_and_Gender.pdf

____________. (2016). OECD/INFE International Survey of Adult Financial Literacy

Competencies. OECD Publishing.

http://www.oecd.org/daf/fin/financial-education/OECD-INFE-International-Survey-of-

Adult-FInancial-LiteracyCompetencies.pdf

Parrotta, J. L., & Johnson, P. J. (1998). The impact of financial attitudes and knowledge on

financial management and satisfaction of recently married individuals. Financial

Counseling and Planning, 9(02), 59-74.

Potrich, A. C. G., Vieira, K. M., Coronel, D. A., & Bender Filho, R. (2015). Financial literacy

in Southern Brazil: Modeling and invariance between genders. Journal of Behavioral

and Experimental Finance, 6, 1-12.

Potrich, A. C. G., Vieira, K. M., & Kirch, G. (2015). Determinantes da Alfabetização

Financeira: análise da influência de variáveis socioeconômicas e demográficas. Revista

Contabilidade & Finanças, 26(69), 362-377.

 40

Savoia, J. R. J, Saito, A. T., & Santana, F. A. (2007). Paradigmas da educação financeira no

Brasil. Revista de Administração Pública, 41(06), 1121-1141.

Schmeiser, M. D., & Seligman, J. S. (2013). Using the Right Yardstick: Assessing Financial

Literacy Measures by Way of Financial Well-Being. Journal of Consumer Affairs,

47(02), 243-262.

Sekita, S. (2011). Financial literacy and retirement planning in Japan. Journal of Pension

Economics and Finance, 10(04), 637-656.

Shanks, T. R. W., & Danziger, S. K. (2010). Anti-Poverty Programs and Policies for Families

and Children. In Jeffrey M. Jenson and Mark W. Fraser, eds., Social Policy for

Children and Families: A Risk and Resilience Perspective: Sage Publications, Inc, pp.

25.

Shockey, S. S. (2002). Low-wealth adults financial literacy. Money management behavior

and associates factors, including critical thinking. Thesis, Utah University, United

States.

Stango, V., & Zinman, J. (2009). Exponential Growth Bias and Household Finance. Journal

of Finance, 64(06), 2807-2849.

United Nations Statistics Division (UNSD). (2013). Gender statistics manual: integrating a

gender perspective into statistics. New York: United Nations, Statistics Division, 2013.

http://unstats.un.org/unsd/genderstatmanual/Sitem ap.ashx

Van Rooij, M. C. J., Lusardi, A., & Alessie, R. J. M. (2011). Financial Literacy and

retirement planning in the Netherlands. Journal of Economic Psychology, 32(04), 593-

608.
 41

Appendix 1
Translated instrument
Financial Attitude (1 = Strongly disagree, 2 = Disagree, 3 = Indifferent, 4 = Agree, 5 = Completely agree)
1 It is important for a family to develop a regular pattern of saving and stick to it.**
2 Families should write financial goals that help them determine priorities in spending.**
3 A written budget is absolutely essential for successful financial management.**
4 It is essential to plan for the possible disability of a family wage earner.**
5 Planning for spending money is essential to successfully managing one's life.**
6 Planning for the future is the best way of getting ahead.**
7 Thinking about where you will be financially in 5 or 10 years is essential to financial success.**
8 Families should really concentrate on the present when managing their finances.
9 Financial planning for retirement is not really necessary for assuring one's security in old age.
10 Having a financial plan makes it difficult to make financial investment decisions.
11 Having a savings plan is not really necessary in today's world in order to meet one's financial needs.
12 Planning is an unnecessary distraction when families are just trying to get by today.
13 Keeping records of financial matters is too time-consuming to worry about.
14 Saving is not really important.
15 As long as one meets monthly payments there is no need to worry about the length of time it will take to
pay off outstanding debts.
Financial Behavior (1 = Never, 2 = Almost never, 3 = Sometimes, 4 = Almost always, 5 = Always)
16 I take notes and control my personal expenses (e.g., expense and revenue spreadsheet).**
17 I compare prices when buying something.
18 I save some of the money I get each month for a future need.**
19 I have a plan for expenses/budget.**
20 I can identify how much I pay when using credit.**
21 I pay my bills without delay.**
22 I save monthly.**
23 I analyze my financial situation before a major purchase.**
24 I always pay my credit cards on time to avoid extra charges.
25 I save regularly to achieve financial targets in the long term.**
26 I save more when I get a pay rise.**
27 I have a financial reserve at least three times my monthly earnings, which can be used in unexpected
circumstances.**
28 In the last 12 months, I have been able to save money.**
Financial Knowledge (Multiple-choice questions)
Q29. Imagine you have R$100.00 in a savings account and the interest rate is 10% per year. After 5
years, how much money will you have in this account?
* More than R$150.00. Less than R$150.00.
Exactly R$150.00. Do not know.
Q30. Imagine the tax rate applied to your savings account is 6% per year and the inflation rate is 10%
per year. After one year, how much will you be able to buy with the money from this account?
Basic Financial Knowledge

More than today. * Less than today.

Exactly the same. Do not know.
Q35. Imagine you take a loan of R$10,000.00 to be paid after one year and the total loan amount with
interest is R$10,600.00. The interest rate you will pay this loan is:
0.3%. * 6%.
0.6%. Do not know.
Q36. Imagine you saw the same television at two different stores at the initial price of R$1000.00.
Store A offers a discount of R$150.00, while the Store B offers a discount of 10%. What is the best
alternative?
* Buy in Store A (discount of R$150.00). Do not know.
Buy in Store B (discount of 10%).
Q37. Imagine five friends receive a donation of R$1,000.00 and they must split the money equally
between them. How much each will get?
100. 5,000.
* 200. Do not know.
 42

Q31. Considering a long time period, (e.g., 10 years), which asset described below normally gives the
highest rate of return?
Savings accounts. Public securities.
Advanced Financial Knowledge

* Stocks. Do not know.

Q32. Normally, which asset shows the highest oscillations over time?
Savings accounts. Public securities.
* Stocks. Do not know.
Q33. When an investor diversifies, his investments are divided among different assets. The risk of losing
money then:
Increases. Remains the same.
* Decreases. Do not know.
Q34. A loan lasting 15 years typically requires higher monthly payments than a 30-year credit, but the
total interest paid at the end of the loan will be less. This affirmation is:
* True. False. Do not know.
Q38. An investment with a high rate of return will have high risk. This affirmation is:
* True. False. Do not know.
Note: * Correct answer. ** Validated questions.