Preliminary draft – please do not cite or distribute without authors permission
The impact of cognitive abilities on financial behavior and well-being
Thérèse Lind, Kenny Skagerlund, Camilla Strömbäck, Gustav Tinghög and Daniel Västfjäll
Linköping University, Sweden
Abstract
In this study we investigate the effect of financial literacy, numeracy and cognitive reflection
on both financial behavior and financial well-being. The findings indicate that when each
cognitive ability, financial literacy, numeric- and cognitive reflective ability, is evaluated
separately higher levels of that ability has a positive effect on financial behavior as well as
on financial well-being. Financial literacy is the most robust predictor and affect financial
behavior and well-being in the subsequent regression analyses when several traits and
demographics are controlled for. These results have important ramifications for
policymakers and whether or not to invest in financial literacy education to enhance
individual financial decision-making skills and individual well-being.
1. Introduction
For many individuals financial matters are synonymous with stress and anxiety. Even though
most of us think that financial matters are tedious there is a great heterogeneity in our
financial behavior. Some of us display quite sound financial behavior while others fail and
therefore face serious consequences. For example, in Sweden one out of seven would not be
able to handle an unforeseen expense of 80001 SEK (SCB). In the US nearly half of the adult
population reports being ill-prepared for a financial disruption and would struggle to cover
emergency expenses of $400 (FRB). Further, studies show that people tend to save too little
for retirement (Lusardi 1999; Munnell et al. 2014), do not set aside money for an emergency
fund (Babiarz & Robb 2014), overspend and experience distress by thinking about their
financial situation. How can we explain this variability in sound financial behavior and wellbeing?
In this paper we investigate if there are some cognitive abilities, such as financial literacy,
that individuals can improve in order to make more informed choices and therefore display
1
Approximately 900 USD
1
Preliminary draft – please do not cite or distribute without authors permission
better financial behavior. Moreover, do individuals with higher levels of financial literacy
feel more secure in their financial situation? In a society where individuals to a great extent
are in charge of their own retirement savings and have the opportunity to invest in complex
financial products these questions are of great importance to policymakers. If individuals do
not have the adequate abilities to handle their financial matters it can have serious
consequences both at an individual level and on an aggregate level. Potential problems could
include low levels of retirement savings, further increases in debt restructuring and greater
financial inequalities between socioeconomic groups.
Previous research has suggested that financial literacy is a key cognitive ability2 that can be
targeted to increase sound financial behavior. Basic financial literacy is the ability to
understand how money works and the set of skills and knowledge that allows people to make
sound financial decisions. This includes a basic understanding of concepts such as
compounded interest rate, inflation and diversification (Lusardi & Mitchell 2007). Among the
financial behaviors previously studied in relation to financial literacy are retirement planning
(Van Rooij et al., 2011b; Almenberg & Säve-Söderbergh, 2011; Lusardi & Mitchell, 2007),
saving for an emergency fund (Babiarz & Robb 2014), stock market participation (Van Rooij
et al. 2011a), credit card-debt (Norvilitis et al. 2006) and overall household wealth (Van
Rooij et al. 2012).
Van Rooij et al. (2011b) show that financial literacy has a causal positive effect on retirement
preparedness although retirement preparedness did not increase in the aftermath the financial
crisis. Almenberg and Säve-Söderbergh (2011) also investigated the link between financial
literacy and retirement preparedness. The study show large differences in financial literacy
between individuals with different educational attainment as well as between men and
women. Individuals who reported that they had tried to plan for retirement had higher levels
of financial literacy (Almenberg & Säve-Söderbergh, 2011).
Van Rooij et al. (2011a) considers financial literacy in relation to stock market participation.
The results indicate that individuals who are more financially literate are likelier to invest in
stocks controlling for demographics and other relevant aspects such as education in
Borghans et al. (2008) pointed out that the usage of the words ‘cognitive and non-cognitive factors’ can be
confusing since few abilities are devoid of cognition. Cognitive abilities are often measured using IQ tests or
other tests constructed to measure complex thinking (Parise & Peijnenburg, 2016). In this paper cognitive
factors are factors measured by some kind of knowledge or performance test, while non-cognitive factors are
self-reported measures of behavior.
2
2
Preliminary draft – please do not cite or distribute without authors permission
economics, daily use of economics and risk aversion. Babiarz and Robb (2014) show that
individuals with higher levels of financial literacy has a greater probability of having an
emergency fund.
A natural way of extending the analysis is by considering concepts close to financial literacy.
For example, it could be argued that the concept of compound interest rate is more of a
numeric concept than a financial concept. Therefore it is close at hand to extend the
investigation of financial behavior to numeric ability. In previous work numeracy has been
defined as the ability to process basic probabilities and numerical concepts (Peters et al.
2006). Statistical numeracy can be said to be a sub-set of numeracy, this sub-set is of
particular importance in order to deal with decisions connected to risk. The statistical feature
of numeracy is therefore key to understanding and evaluating, for example, financial
information. (Peters et al. 2006).
By matching measures of numeracy with administrative records with detailed information on
mortgage payments Gerardi et al. (2013) has shown that greater numeric ability is negatively
associated with the propensity to default on one’s mortgage. Grafeo et al. (2015) show that
numeric skill and cognitive reflection predicts the use of a more thorough decision process
resulting in a better consumer choice. Cognitive reflection is the ability to resist the
intuitively incorrect answer suggested by our intuition in favor for the more effortful correct
answer. Previous literature has shown that in relation to financial matters individuals are
especially prone to certain behavioral biases. For example over-investing in familiar stocks or
being susceptible to various framing effects (Huberman, 2001; Byrne, 2005). One aspect that
has been shown to predict less susceptibility to such biases is cognitive reflection. It is
therefore of interest to see if this ability has a direct effect on financial behavior.
Although important, individuals’ financial behavior and what affects it is ones perceived
security about this behavior should not be neglected. Some individuals are more prone to
worrying about financial matters than others affecting their well-being. In this paper we
define financial well-being as how confident or anxious an individual feel towards her
personal finances. The aspect of financial well-being should be of great importance to
policymakers since ones perceived financial situation is a strong predictor of overall life
satisfaction (Melin et al., 2003). Prawits et al. (2006) point out that objective measures have
been used to predict individuals’ perception about their financial situation but such indicators
do not measure the depth of one’s feelings to a certain situation. Further, Johnson and
3
Preliminary draft – please do not cite or distribute without authors permission
Krueger (2006) show that there is an independence between ones perceived financial
situation and actual wealth.
Since there is a missing link in the current literature between financial behavior and financial
well-being it is important to investigate if it is the same explanatory factors that determine
financial behavior that also affect financial well-being. In this paper we do this by examining
both objective and self-reported knowledge in relation to a wide set of financial behaviors.
We use a broad financial management scale in order to better understand the individual
heterogeneity and get a fuller view of individual financial behavior compared to previous
studies. Further, we analyze if the same explanatory variables determine both financial wellbeing and financial behavior. We also control for non-cognitive traits, in particular selfcontrol, in order to avoid the problem of omission bias present in the literature on financial
behavior (Fernandes et al. 2014).
1.1 Hypotheses
Based on results from previous studies we would expect that individuals that score higher on
the financial literacy scale will show a more desirable financial behavior. In other words, the
knowledge based financial literacy translates into a more desirable financial behavior. We
would also except that self-reported financial literacy will have a positive correlation with
measured financial literacy and will also indicate better financial behavior. With regards to
financial well-being our predication is that higher levels of financial literacy, both objective
and subjective will have a positive effect on financial security and a negative effect on
financial anxiety. Whether or not CRT skills or numeracy skills actually affects financial
behavior or financial well-being is much less clear in the literature. However we expect that if
there is an effect more numerate individuals will show more desirable financial behavior. For
example, Sinayev and Peters (2015) has shown that numeric ability is a robust predictor of
superior decision making, although their study did not include financial decisions in
particular. We also expect that higher CRT-score will indicate better financial behavior. This
would be in line with work by Graffeo et al. (2015).
4
Preliminary draft – please do not cite or distribute without authors permission
2. Method and data
2.1 Sample and survey
We surveyed 2063 Swedish adults during April and May of 2016. The survey was distributed
via a web link so the participants could respond at home. Participants were paid for their
participation. The average age in the sample is 49.2 years, 50.8% of the participants are
women and 28.4% of the sample has at least a bachelor’s degree. Of the sample 14.6 %
reported having a monthly income (before tax, at the household level) below 15 000 SEK and
14.4% reported having a monthly income exceeding 55 000 SEK. The average number of
persons per household was 2.2. The survey collected information about financial behavior,
financial literacy, cognitive reflective ability, numeracy and financial well-being among
others.
2.2 Dependent variables
2.2.1 Financial behavior
Financial behavior is measured with the Financial Management Behavior Scale (FMBS)
developed by Dew and Xiao (2011), the scale consists of twelve items. The participants were
asked to indicate, on a scale from one to five, how often they had engaged in a number of
activities over the past six months. The items are displayed in table 1.
Table 1. Financial Management Behavior Scale (FMBS)
1
2
3
4
5
6
7
8
9
10
11
12
Comparison shopped when purchasing a product or service
Paid all your bills on time
Kept a written or electronic record of your monthly expenses
Stayed within your budget or spending plan
Paid off credit card balance in full each month
Maxed out the limit on one or more credit cards
Made only minimum payments on a loan
Began or maintained an emergency savings fund
Saved money from every paycheck
Saved for a long term goal such as a car, education, home, etc.
Contributed money to a retirement account
Bought bonds, stocks, or mutual funds
A higher score indicates more desirable financial behavior, items six and seven are reversed
to calculate the final score. The scale is compelling since it considers basic and yet very
important financial behaviors. These behaviors are needed in order to ensure a sound
5
Preliminary draft – please do not cite or distribute without authors permission
financial situation, both in the short-run (payed all bills on time) and long-run (saving for
retirement or other long-term goal). To our knowledge the FMBS has not previously been
tested on a Swedish population.
2.2.2 Financial well-being
We measure participants’ level of financial well-being with two separate scales, financial
anxiety and financial security. The financial anxiety scale was developed by Fünfgeld and
Wang (2009) and the items are displayed in table 2a. In the second scale we wanted to see
how confident respondents are in their current financial situation but also how they feel
towards their future financial situation, this resulted in the three items stated in table 2b.
Respondents were asked to indicate on a five point Likert scale how well each statement
corresponds to their own situation. Five indicating that the respondent agrees completely with
the statement and one indicating that the respondent does not at all agree.
Table 2a. Financial Anxiety (FA)
1
2
3
4
I get unsure by the lingo of financial experts
I am anxious about financial and money affairs
I tend to postpone financial decisions
After making a decision, I am anxious whether I was right or wrong
Table 2b. Financial Security (FS)
1
I feel secure in my current financial situation
2
I feel confident about my financial future
I feel confident about having enough money to support myself in retirement,
no matter how long I live
3
A factor analysis confirmed our hypothesis that the two scales measure two underlying
constructs. Individuals can feel quite comfortable with their financial situation and yet worry
about over financial matters and decisions. Thus, we include these two subcomponents into
the more general notion of financial well-being.
2.3 Independent variables
2.3.1 Financial literacy, numeracy and cognitive reflection
In order to measure financial literacy we included four knowledge based questions as well as
one question where participants were asked to self-report their degree of economic and
6
Preliminary draft – please do not cite or distribute without authors permission
financial knowledge. The first three questions test very basic financial knowledge, the first on
is a simple numeracy/interest rate question, the second one is tests the subject’s knowledge
about inflation. The third question asked about the concept of diversification. The last
question is a bit more advanced testing the participant’s knowledge about the relationship
between bond prices and interest rate. All questions have been used in previous studies, for
example by Van Rooij, Lusardi, & Alessie (2012).We chose to only include the four
questions in order to make the survey of reasonable length, and many financial literacy
studies have previously used only around three or four questions (Almenberg & SäveSöderbergh, 2011; Van Rooij, Lusardi & Alessie 2011b; Babiarz & Robb, 2014).
Numeracy was measured measured using a combined score of the three items from Schwartz
et al. (1997) and with the Berlin Numeracy Test (BNT), developed by Cokely et al. (2012)
and validated in Swedish by Lindskog et al. (2015). This scale was chosen since it
specifically test statistical numeracy and risk literacy, both of importance when dealing with
financial issues. The combination of the questions from Schwartz et al., (1997) and the BNT
was used due to the investigation of a representative sample of the general population,
whereas the BNT was primarily developed to assess numeracy in educated samples. The
combined use of these instruments has also been advocated by Cokeley et al. (2012).
The numeracy task consisted of 7 questions were all participants answered the first four
questions. Depending on the participants answer they moved on according to the path
presented in figure 1. All subject were presented with at least five questions.
If participants answered question 4 correctly they moved on to question 5b and if answered
question 4 incorrectly they moved on to 5a. After answering question 5a the participants were
done with the numeracy task. If participants answered 5b correctly they moved on to question
6, otherwise they were done with the task. One point was rewarded to each correct answer,
except for a correct answer to the fourth question that rewarded the participant with two
points. Consequently the numeracy score varies between zero and seven.
We chose to measure cognitive reflection by means of Fredericks (2005) three item scale,
called the Cognitive Reflection Test (CRT). Consult the appendix to see all measures and
items used in the survey. The CRT-score is simply the number of correct answers given to the
three CRT items developed by Frederick (2005)
7
Preliminary draft – please do not cite or distribute without authors permission
3. Empirical Results
3.1 Descriptive evidence on financial behavior, financial well-being, financial literacy,
numeracy and CRT
As mentioned before the dependent variable is a measure developed by Dew and Xiao
(2011), table 3 shows the distribution of each item as well as the average measure used
further in the analysis. N/A indicates that the participants were able to choose the option not
applicable. This option was excluded in as many items as possible in order to encourage
active participation.
Table 3. Financial Management Behavior Scale (FMBS)
FMBS
1
Comparison shopped when purchasing a product or service
2
Paid all your bills on time
3
Kept a written or electronic record of your monthly expenses
4
Stayed within your budget or spending plan
5
Paid off credit card balance in full each month
6
Maxed out the limit on one or more credit cards
7
Made only minimum payments on a loan
8
Began or maintained an emergency savings fund
9
Saved money from every paycheck
10 Saved for a long term goal such as a car, education, home, etc.
11 Contributed money to a retirement account
12 Bought bonds, stocks, or mutual funds
FMBS Average
Mean St.dev. Range
3.86 1.00
1-5
4.56 0.84
1-5
3.54 1.34
1-5
2.98 1.50 1-5, N/A
3.20 1.87 1-5, N/A
1.60 1.03 1-5, N/A
2.50 1.34
1-5
3.23 1.41
1-5
3.54 1.40
1-5
3.10 1.41
1-5
2.90 1.56
1-5
2.53 1.45
1-5
3.44 0.65
1.5-5
*The not applicable answers where coded as 1. 1 This item was reversed when calculating the average level.
The average level on the FMBS is 3.44, indicating that participants on average engaged in the
listed activities quite regularly over the past six months. The question with the highest mean
is the second question indicating that a majority in the sample paid all their bills on time
every month.
Since the FMBS has not previously been tested in Sweden it is interesting to see the
distribution to each item, in table 4 these results are presented. For example 72 % of the
sample reported having paid all their bills on time during the last six months. 25 %
contributed to a retirement account every month during the last six month.
8
Preliminary draft – please do not cite or distribute without authors permission
3.2 Financial literacy
Table 4. Distribution of participants answers to each FMBS item, in percent. 1 = never engaged in this activity
during the last six month and 5 = always engaged in this activity during the last six month.
FMBS
1
2
3
4
5
Comparison shopped when purchasing a product or service
2.57
5.62
25.74
35.19
30.88
Paid all your bills on time
1.26
2.38
7.90
16.34
72.13
Kept a written or electronic record of your monthly expenses
9.50
14.54
22.44
19.53
33.98
Stayed within your budget or spending plan
30.15
5.09
19.68
26.85
18.23
Paid off credit card balance in full each month
39.51
1.11
6.01
6.83
46.53
Maxed out the limit on one or more credit cards
68.69
12.65
11.29
4.75
2.62
Made only minimum payments on a loan
34.37
13.57
31.17
9.60
11.29
Began or maintained an emergency savings fund
17.69
12.02
26.08
18.47
25.74
Saved money from every paycheck
11.68
12.55
22.73
15.75
37.28
Saved for a long term goal such as a car, education, home, etc.
18.37
16.09
25.98
16.53
23.02
Contributed money to a retirement account
30.25
12.65
18.81
13.72
24.58
Bought bonds, stocks, or mutual funds
37.42
13.62
21.62
13.14
14.20
The first column of table 5 shows the percent of correct answers to each individual financial
literacy item. (The exact formulation can be found in the appendix). The second column
shows the percent of respondents who chose the alternative do not know. As means of
comparison we have also included the distributions from two previous studies (Almenberg &
Säve-Söderbergh, 2011; Van Rooij et al., 2012) that used approximately the same financial
literacy questions, this is shown in the subsequent columns. The first question in our study is
not exactly comparable to the results by Almenberg & Säve-Söderbergh (2011) since they
asked their participants to make their own calculation, whereas our participants faced a
multiple-choice question. In this respect our results would be more comparable to the results
by Van Rooij et al. (2012). Also in the risk diversification question the results are more in
line with the result presented in Van Rooij et al. The fourth and most difficult financial
literacy question that was included in this study was related to the relationship between bond
prices and interest rate, here only 15.7 % of the participants chose the correct option. The
same question was included in the paper by Van Rooij et al were respondents answered
9
Preliminary draft – please do not cite or distribute without authors permission
correctly 24.6 % of the time.
Table 5. Comparison between number of correct financial literacy questions, percent
Present study
Financial Literacy questions
Correct
A&S 2011
VR, L&A 2012
DK* Correct DK* Correct
DK*
Interest rate /Numeracy
77.2
9.4
35.2
15.6
90.8
3.7
Inflation
63.1
21.4
59.5
16.5
82.6
8.5
Diversification
51.0
41.3
68.4
18.4
48.2
26.6
Interest rate-bond price relationship
15.7
36.9
-
-
24.6
37.5
Note:*DK stands for do not know. In both A&S 2011 and VR, L&A 2012 the option refuse to answer was
include, since no one used this option in A&S 2011 we chose to exclude it.
On average the participants answered 2.1 financial knowledge questions correctly.
Approximately 10 % of the sample answered all four financial literacy questions correctly.
Participants were also asked to self-report their level of financial literacy. The scale ranged
between one and seven. The mean self-reported financial literacy was 4.32 and the median
was 4, the correlation between the knowledge based score and the self-reported measure was
r =.34.
10
Preliminary draft – please do not cite or distribute without authors permission
Graph 1.
In order to understand how financial literacy relates to financial behavior we divide the
sample into two subgroups, those with low levels of financial knowledge (a score lower than
three) and those with high level of financial knowledge (a score equal to or higher than
three). Approximately 60 % of the sample has low financial knowledge according to this
definition. In order to see if these two groups differ with respect to their financial behavior
we do a simple t-test which shows that the difference is statistically significant (pvalue=0.00). The results can be seen in the graph 1.
The participants with low levels of financial literacy has a mean score of approximately 3.30
on the FMBS and participants with a high financial literacy score has an average of 3.66 on
the FMBS.
3.3 Numeracy
In table 6 the percent of correct answers to each numeracy item are presented, for example 67
% of the sample where able to correctly estimate how many times out of a 1000 a fair coin
would show tails. Only 23% of the sample answered the last “common” question (question 4)
correctly. See appendix for exact formulation of the questions.
11
Preliminary draft – please do not cite or distribute without authors permission
Table 6. Numeracy task, percent of correct answers
11
21
31
41
5a2
5b3
64
Fair coin toss
Number of people winning 10 dollar prize
Percent of winning tickets
Men in the choir
Five-sided die
Loaded die
Red poisonous mushroom
67.14
62.24
49.88
22.73
25.85
32.46
40.99
1
Shown to whole sample 2Shown to those who answered 4 incorrectly
3
Shown to those who answered 4 correctly 4Shown to those who answered 5b correctly
Points were rewarded to each correct answer as mentioned above, the distribution of the total
scores can be seen in table 7. For example, 16.9% of the sample answered correctly to 3 of
the numeracy questions, although this measure does not indicate which questions the
participant’s answered correctly. On average participants answered 2.6 questions correctly.
Table 7. Numeracy score
distribution, in percent
0
1
2
3
4
5
6
7
15.41
19.34
17.26
16.87
14.20
9.99
4.02
2.91
We undertook the same procedure with the numeracy score as with the financial literacy
score, that is, we divided the participants into two groups with respect to their level of
numeracy and compare them in terms of financial behavior. Here we classify high numeracy
as having a score equal to or above five, with this definition approximately 83 % of the
sample has low numerical skills. The results are shown in graph 2, there is a small but
12
Preliminary draft – please do not cite or distribute without authors permission
statistically significant (p-value=0.01) difference between the groups. Those with high
numeric skills has somewhat of a higher mean on the FMBS than those with low numeric
skills, 3.53 compared to 3.43.
3.4 CRT
The CRT score is simply the number of correct answers given to the three CRT items
developed by Frederick (2005). In the present sample 48.2 % of the participants answered
incorrectly to all thee CRT items. 24 % answered one question correctly, 17.8 % answered
two questions correctly and lastly 9.94 % answered all three CRT items correctly. The sample
mean CRT score is 0.89.
Once more we divide the sample into two groups, categorizing those with two or three correct
answers as having high reflective ability whereas those with zero or one correct answer has a
low reflective ability. The results are shown in graph 3. There is a statistically significant (pvalue=0.00) difference with regards to financial behavior. Those with higher reflective ability
display somewhat more desirable financial behavior than those with lower reflective ability.
On the FMBS scale the different means are 3.53 (for high ability) and 3.41 (for low ability).
13
Preliminary draft – please do not cite or distribute without authors permission
To sum up, when investigating each skill on its own with a difference in means approach
there seems to be a difference with regards to financial behavior with respect to each
discussed ability. The signs of these differences are in line with prior studies, that is
individuals with lower abilities display less desirable financial behavior.
3.5 Determinants of financial behavior - a multivariate analysis
Firstly we predict financial behavior in three regressions using each explanatory ability
(financial literacy, numeracy and cognitive reflection) separately. We control for, age,
education, gender, income and whether the respondent was born in a non-European country.
The results are presented in table 8. Even after controlling for demographics the same pattern
as in the difference in means analysis emerge. For financial literacy the estimate indicates
that one more correct answer to the financial knowledge questions increases the average
FMBS value by 0.12 points.
In table 9 we report the estimates of a different specifications on financial literacy and
successively add the explanatory variables into one model. In the first and second column
numeracy and cognitive reflection enters the model respectively, while measuring financial
literacy as the number of correct answers (as in column 1 of table 5). In model (3) all three
explanatory abilities enter the model simultaneously but neither numeracy nor the cognitive
reflection are statistically significant and thus do not have a significant impact on financial
behavior. In the fourth specification we add a measure of self-reported financial knowledge,
this measure is statistically significant and reduces the estimate of the financial literacy score
somewhat. In column (5) financial literacy enters the regression as a dummy variable, 1 if
that individual answered correctly to all financial knowledge questions 0 otherwise.
Numeracy and cognitive reflection also enters the model as dummy variables.
Noteworthy is that no matter the specification financial literacy is statistically significant,
although the effect size is quite small. What is also important to note is the results in model
(3), when financial literacy is included as an explanatory variable numeracy and CRT scores
are not statistically significant. This is a very interesting results since previous studies has
shown numeric ability to be strongly associated with better decision making.
Self-assessed financial literacy (added in model (4) in table 9) clearly matter for one’s
financial behavior. When this variable is added to the model the magnitude of the financial
14
Preliminary draft – please do not cite or distribute without authors permission
literacy estimate decreases somewhat (even though it is still statistically significant) and selfassed literacy has a positive and significant effect on financial behavior.
Table 8. Financial Management Behavior Scale (FMBS)
VARIABLES
(1)
Financial literacy, number of correct answers
OLS
(2)
0.115***
(0.012)
Numeracy score
0.022***
(0.007)
CRT score
0.029**
(0.013)
Self-control
Female
Born in a non-European country
Age
University education
Upper secondary education
Income 15 000-24 999 SEK
Income 25 000-34 999 SEK
Income 35 000-44 999 SEK
Income 45 000-54 999 SEK
Income 55 000-64 999 SEK
Income 65 000-74 999 SEK
Income>75 000 SEK
Observations
R-squared
(3)
0.178***
(0.019)
0.050*
(0.027)
-0.029
(0.077)
0.004***
(0.001)
0.050
(0.035)
0.008
(0.031)
0.091**
(0.046)
0.253***
(0.045)
0.357***
(0.049)
0.394***
(0.055)
0.472***
(0.055)
0.493***
(0.064)
0.500***
(0.055)
2,062
0.222
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
0.183***
(0.019)
-0.007
(0.027)
-0.028
(0.083)
0.005***
(0.001)
0.083**
(0.035)
0.006
(0.032)
0.103**
(0.047)
0.260***
(0.046)
0.387***
(0.050)
0.418***
(0.057)
0.494***
(0.056)
0.540***
(0.065)
0.529***
(0.056)
0.184***
(0.019)
-0.016
(0.027)
-0.040
(0.083)
0.005***
(0.001)
0.088**
(0.035)
0.006
(0.032)
0.106**
(0.047)
0.264***
(0.046)
0.393***
(0.050)
0.423***
(0.056)
0.502***
(0.056)
0.545***
(0.065)
0.537***
(0.057)
2,062
0.191
2,062
0.190
15
Preliminary draft – please do not cite or distribute without authors permission
Table 9. Financial Management Behavior Scale (FMBS)
OLS
VARIABLES
(1)
(2)
Financial literacy, number of correct answers
Numeracy score
CRT score
0.127***
(0.014)
-0.012
(0.009)
-0.007
(0.016)
0.081***
(0.013)
-0.010
(0.009)
0.005
(0.015)
0.150***
(0.012)
Self-reported financial knowledge
Financial literacy dummy, all correct
Numeracy Score>=5
CRT Score>=2
Self-control
Female
Born in a non-European country
Age
University education
Upper secondary education
Income 15 000-24 999 SEK
Income 25 000-34 999 SEK
Income 35 000-44 999 SEK
Income 45 000-54 999 SEK
Income 55 000-64 999 SEK
Income 65 000-74 999 SEK
Income>75 000 SEK
Observations
R-squared
0.179***
(0.019)
0.044
(0.027)
-0.039
(0.077)
0.003***
(0.001)
0.057
(0.035)
0.008
(0.031)
0.091**
(0.046)
0.256***
(0.045)
0.359***
(0.049)
0.402***
(0.055)
0.478***
(0.055)
0.495***
(0.064)
0.504***
(0.055)
2,062
0.223
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
(3)
0.181***
(0.045)
-0.014
(0.038)
0.004
(0.032)
0.185***
(0.019)
-0.006
(0.027)
-0.050
(0.082)
0.005***
(0.001)
0.087**
(0.035)
0.006
(0.032)
0.106**
(0.047)
0.265***
(0.046)
0.388***
(0.050)
0.430***
(0.056)
0.497***
(0.056)
0.540***
(0.065)
0.528***
(0.057)
0.137***
(0.018)
0.070***
(0.026)
-0.090
(0.075)
0.002***
(0.001)
0.049
(0.033)
0.018
(0.030)
0.048
(0.045)
0.200***
(0.043)
0.293***
(0.047)
0.323***
(0.054)
0.386***
(0.053)
0.419***
(0.062)
0.416***
(0.054)
2,062
0.194
2,062
0.290
16
Preliminary draft – please do not cite or distribute without authors permission
3.7 Determinants of financial well-being - a multivariate analysis
Lastly we investigate if the same abilities that influence financial behavior also affects
financial well-being. As mentioned previously two different dependent variables were
considered in order to understand financial well-being, financial security and financial
anxiety. A difference in mean approach reveal the same pattern as with financial behavior, if
the sample is divided into two groups, those with high and those with low ability there is a
significant difference between the two groups both with regards to financial security and
anxiety. The results of the t-test are presented in table.
Table 13. t-test between high and low levels of financial literacy, self-reported financial
knowledge, numeracy and CRT
Financial
literacy
High
Low
Average financial
security
Average financial
anxiety
Self-perceived financial
knowledge
High
Low
Numeracy
High
Low
CRT
High
Low
3.76
2.94
2.56
3.55
3.32
2.96
3.24
2.94
2.41
2.85
2.58
3.01
2.83
2.71
2.72
2.84
Note: For the pairs stated in bold numbers the double sided t-test between the high ability and the low ability
group show p-values<0.01, t for the pair not marked in bold the double sided t-test have a p-value<0.05
Noteworthy is the large difference between the high and low ability groups with respect to
both actual financial literacy and perceived financial knowledge. Those who answer correctly
to three (out of the four) financial literacy questions has an average financial security of 3.76
compared with 2.94 for those with less than three correct answers. This is a much larger
difference than for behavior.
To see whether these differences persist when controlling for demographics we proceed with
a regression analysis. The OLS estimates are reported in table 14. In the first column we find
that increased financial knowledge indicates higher financial security. The second
specification (column two) shows that increased financial knowledge lowers the respondent’s
level of anxiety. In columns three and four the self-reported measure of financial knowledge
is added. With this addition the statistical significance of the knowledge based measure
disappear in column three. This result is perhaps not to surprising, it is what a person believes
she knows that explains her well-being in that particular matter. It is more surprising that the
17
Preliminary draft – please do not cite or distribute without authors permission
statistical significance of the financial literacy measure persist (at a 5 % significance level) in
column four.
Table 14. Financial well-being
OLS
VARIABLES
Financial literacy, number of
correct answers
Numeracy score
CRT score
Financial Security
(1)
(2)
0.136***
(0.025)
0.032
(0.023)
-0.086***
(0.018)
-0.053***
(0.018)
-0.007
(0.017)
-0.012
(0.027)
-0.003
(0.015)
0.016
(0.026)
0.345***
(0.020)
0.013
(0.012)
-0.006
(0.019)
0.012
(0.012)
-0.015
(0.019)
-0.110***
(0.017)
0.358***
(0.034)
-0.168***
(0.049)
-0.033
(0.151)
0.009***
(0.002)
-0.089
(0.056)
0.056
(0.063)
0.350***
(0.083)
0.700***
(0.084)
0.852***
(0.088)
0.998***
(0.100)
1.176***
(0.091)
1.181***
(0.106)
1.328***
(0.099)
0.263***
(0.031)
-0.109**
(0.045)
-0.153
(0.137)
0.007***
(0.001)
-0.065
(0.052)
0.038
(0.058)
0.252***
(0.077)
0.571***
(0.079)
0.700***
(0.082)
0.816***
(0.095)
0.964***
(0.087)
1.006***
(0.100)
1.125***
(0.095)
-0.368***
(0.025)
0.146***
(0.034)
0.017
(0.100)
-0.002**
(0.001)
0.053
(0.039)
0.012
(0.044)
-0.026
(0.060)
-0.100*
(0.060)
-0.136**
(0.063)
-0.125*
(0.070)
-0.237***
(0.073)
-0.187**
(0.083)
-0.249***
(0.071)
-0.338***
(0.025)
0.127***
(0.034)
0.055
(0.101)
-0.002
(0.001)
0.045
(0.039)
0.018
(0.043)
0.006
(0.060)
-0.059
(0.060)
-0.087
(0.063)
-0.067
(0.068)
-0.169**
(0.072)
-0.131
(0.083)
-0.185***
(0.071)
2,062
0.198
2,062
0.222
Self-reported financial knowledge
Self-control
Female
Born in a non-European country
Age
Upper secondary education
University education
Income 15 000-24 999 SEK
Income 25 000-34 999 SEK
Income 35 000-44 999 SEK
Income 45 000-54 999 SEK
Income 55 000-64 999 SEK
Income 65 000-74 999 SEK
Income>75 000 SEK
Observations
R-squared
Financial Anxiety
(3)
(4)
2,062
2,062
0.290
0.393
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
18
Preliminary draft – please do not cite or distribute without authors permission
4. Discussion
Every day individuals face a plethora of choices, many of them with financial consequences.
Today individuals also face increased exposure to complex financial products and increased
responsibility for retirement savings, this combination put a lot of pressure on the individual
decision maker. A behavior that has been well studied in the past is the individuals choice of
how much to save, although this is a very important decision we argue that many other
factors also determine ones overall financial behavior, this can be decisions spanning from
paying ones bills on time to regularly checking ones expenditures. In this article we fill this
gap by studying a much broader set of financial behaviors than previously considered. We
further investigate if certain abilities, such as financial literacy, numeracy and cognitive
reflection, affect this broad set financial behaviors and also whether or not the same factors
affect financial well-being.
The analyses reveal that when financial literacy, numeracy and cognitive reflection are
considered in isolation they have an effect on financial behavior as well as financial wellbeing. When controlling for demographics and financial literacy, abilities such as numeracy
and cognitive reflection does not influence behavior in a statistically significant way. We
investigate the relationship between the three abilities by instead using cognitive reflection
and numeracy as instruments for financial literacy, this in order to avoid reversed causality
problems.
Actual financial literacy affect the financial behaviors considered in all specifications, OLS
as well as 2SLS, when it comes to financial well-being the self-reported measure has a larger
impact than the objective measure on behavior. This could have serious consequences since
the correlation between actual and perceived financial knowledge is only .34 in the sample. If
individuals feel comfortable and not very anxious about their financial situation due to that
they believe they have adequate financial knowledge when in fact their knowledge level is
quite low their financial decisions will be biased.
This study, and many others, confirm that financial literacy matters for financial behavior but
another well documented finding is that financial literacy education seems to have very little
effect on teaching people about everyday financial concepts (Fernandes et al. 2014). One
explanation for this might be the relatively small effect that financial literacy has on behavior.
19
Preliminary draft – please do not cite or distribute without authors permission
Financial literacy is usually seen as an investment in human capital with is both costly and
time-consuming (Jappelli & Padula, 2013). Then it must be that for many individuals the cost
of acquiring financial literacy is greater than the benefit it generates.
20
Preliminary draft – please do not cite or distribute without authors permission
References
Almenberg, J., & Säve-Söderbergh, J. (2011). Financial literacy and retirement planning in
Sweden. Journal of Pension Economics and Finance, 10(04), 585-598.
Atkinson, A., & Messy, F. A. (2011). Assessing financial literacy in 12 countries: an
OECD/INFE international pilot exercise. Journal of Pension Economics and Finance, 10(04),
657-665.
Babiarz, P., & Robb, C. A. (2014). Financial literacy and emergency saving. Journal of
Family and Economic Issues, 35(1), 40-50.
Borghans, L., Duckworth, A. L., Heckman, J. J., & Ter Weel, B. (2008). The economics and
psychology of personality traits. Journal of human Resources, 43(4), 972-1059.
Brown, M., & Graf, R. (2012). Financial literacy, household investment and household debt:
Evidence from Switzerland. Working Papers on Finance, 1301.
Bureau, C. F. P. (2015). Financial well-being: The goal of financial education. Report, Iowa
City, IA: Consumer Financial Protection Bureau.
Bucher-Koenen, T., & Ziegelmeyer, M. (2011). Who lost the most? Financial literacy,
cognitive abilities, and the financial crisis.
Byrne, K. (2005). How do consumers evaluate risk in financial products?. Journal of
Financial Services Marketing, 10(1), 21-36.
Cokely, E. T., Galesic, M., Schulz, E., Ghazal, S., & Garcia-Retamero, R. (2012). Measuring
risk literacy: The Berlin numeracy test. Judgment and Decision Making, 7(1), 25.
Dew, J., & Xiao, J. J. (2011). The financial management behavior scale: Development and
validation. Journal of Financial Counseling and Planning, 22(1), 43.
Fernandes, D., Lynch Jr, J. G., & Netemeyer, R. G. (2014). Financial literacy, financial
education, and downstream financial behaviors. Management Science, 60(8), 1861-1883.
Frederick, S. (2005). Cognitive reflection and decision making. The Journal of Economic
Perspectives, 19(4), 25-42.
FRB, Survey of Household Economics and Decisionmaking 2016
Fünfgeld, B., & Wang, M. (2009). Attitudes and behaviour in everyday finance: evidence
from Switzerland. International Journal of Bank Marketing, 27(2), 108-128.
Graffeo, M., Polonio, L., & Bonini, N. (2015). Individual differences in competent consumer
choice: the role of cognitive reflection and numeracy skills. Frontiers in psychology, 6.
Huberman, G. (2001). Familiarity breeds investment. Review of financial Studies, 14(3), 659680.
Hung, A., Parker, A. M., & Yoong, J. (2009). Defining and measuring financial literacy.
Jappelli, T., & Padula, M. (2013). Investment in financial literacy and saving
decisions. Journal of Banking & Finance, 37(8), 2779-2792.
21
Preliminary draft – please do not cite or distribute without authors permission
Johnson, W., & Krueger, R. F. (2006). How money buys happiness: genetic and
environmental processes linking finances and life satisfaction. Journal of personality and
social psychology, 90(4), 680.
Kahneman, D., & Frederick, S. (2002). Representativeness revisited: Attribute substitution in
intuitive judgment. Heuristics and biases: The psychology of intuitive judgment, 49.
Lindskog, M., Kerimi, N., Winman, A., & Juslin, P. (2015). A Swedish validation of the
Berlin Numeracy Test. Scandinavian journal of psychology, 56(2), 132-139.
Lusardi, A., & Mitchell, O. S. (2007). Financial literacy and retirement planning: New
evidence from the Rand American Life Panel. Michigan Retirement Research Center
Research Paper No. WP, 157.
Lusardi, A., & Mitchell, O. S. (2011). Financial literacy around the world: an overview.
Journal of Pension Economics and Finance, 10(04), 497-508.
Mandell, L., & Klein, L. S. (2009). The impact of financial literacy education on subsequent
financial behavior. Journal of Financial Counseling and Planning,20(1).
Mani, A., Mullainathan, S., Shafir, E., & Zhao, J. (2013). Poverty impedes cognitive
function. science, 341(6149), 976-980.
Melin, R., Fugl-Meyer, K. S., & Fugl-Meyer, A. R. (2003). Life satisfaction in 18-to 64-yearold Swedes: in relation to education, employment situation, health and physical activity.
Journal of rehabilitation medicine, 35(2), 84-90.
Munnell, A. H., Webb, A., & Hou, W. (2014). How much should people save?. Center for
Retirement Research at Boston College, July 2014.
Norvilitis, J. M., Merwin, M. M., Osberg, T. M., Roehling, P. V., Young, P., & Kamas, M.
M. (2006). Personality factors, money attitudes, financial knowledge, and credit‐card debt in
college students1. Journal of Applied Social Psychology, 36(6), 1395-1413.
Oechssler, J., Roider, A., & Schmitz, P. W. (2009). Cognitive abilities and behavioral biases.
Journal of Economic Behavior & Organization, 72(1), 147-152.
Parise, G., & Peijnenburg, K. (2016). Noncognitive Abilities and Financial Distress.
Peters, E., Västfjäll, D., Slovic, P., Mertz, C. K., Mazzocco, K., & Dickert, S. (2006).
Numeracy and decision making. Psychological science, 17(5), 407-413.
Prawitz, A. D., Garman, E. T., Sorhaindo, B., O'Neill, B., Kim, J., & Drentea, P. (2006).
InCharge financial distress/financial well-being scale: Development, administration, and
score interpretation. Journal of Financial Counseling and Planning, 17(1).
SCB, Undersökningarna av levnadsförhållanden (ULF/SILC).
Schwartz, L. M., Woloshin, S., Black, W. C., & Welch, H. G. (1997). The role of numeracy
in understanding the benefit of screening mammography. Annals of internal medicine,
127(11), 966-972.
22
Preliminary draft – please do not cite or distribute without authors permission
Sinayev, A., & Peters, E. (2015). Cognitive reflection vs. calculation in decision making.
Frontiers in psychology, 6, 532.
Van Rooij, M., Lusardi, A., & Alessie, R. (2011a). Financial literacy and stock market
participation. Journal of Financial Economics, 101(2), 449-472.
Van Rooij, M. C., Lusardi, A., & Alessie, R. J. (2011b). Financial literacy and retirement
planning in the Netherlands. Journal of Economic Psychology, 32(4), 593-608.
Van Rooij, M. C., Lusardi, A., & Alessie, R. J. (2012). Financial literacy, retirement planning
and household wealth. The Economic Journal, 122(560), 449-478.
Welsh, M., Burns, N., & Delfabbro, P. (2013). The Cognitive Reflection Test: how much
more than numerical ability. In Proceedings of the 35th Annual Conference of the Cognitive
Science Society (pp. 1587-1592). Austin, TX: Cognitive Science Society.
Xu, L., & Zia, B. (2012). Financial literacy around the world: an overview of the evidence
with practical suggestions for the way forward. World Bank Policy Research Working Paper,
(6107).
23
Preliminary draft – please do not cite or distribute without authors permission
Appendices
A1
Sample means and standard deviations
Obs
Mean
Std. Dev.
Min
Max
Average FMBS
2063 3.444296 .6479011
1.5
5
Financial literacy score
2063 2.070286 1.16546
0
4
CRT score
2063 .8948134 1.023055
0
3
Numeracy score
2063 2.556956 1.872909
0
7
Financial literacy questions and choice alternatives
1. Suppose you had 100 kr in a savings account and the interest rate was 2% per year.
After 5 years, how much do you think you would have in the account if you left the money to
grow? Possible answers: More than 102 kr, exactly 102 kr, less than 102 kr, do not know
2. Imagine that the interest rate on your savings account was 1% per year and inflation was
2% per year. After 1 year, would you be able to buy more than, exactly the same as or less
than today with the money in this account? Possible answers: More than today, exactly the
same as today, less than today do not know
3. Buying a single company’s tock usually provides a safer return than a stock mutual fund.
Possible answers: True, false, do not know
4. If interest rates rise, what will typically happen to bond prices? Possible answers: They
will rise, they will fall, they will stay the same, there is no relationship between bond prices
and the interest rates, do not know
Numeracy questions
1. Imagine that we throw a fair coin 1000 times. How many times do you think the coin will
show tails? _____________times out of a 1 000.
24
Preliminary draft – please do not cite or distribute without authors permission
2. In a small American lottery the chance of winning 10 dollars is 1 %. What is your best
guess about how many people will win the 10 dollar prize if 1000 people each by a single
ticket. ___________people out of 1000.
3. In another lottery the chance of winning a car is 1 in 1000. What percent of tickets win a
car? __________%
4. Out of 1,000 people in a small town 500 are members of a choir. Out of these 500
members in the choir 100 are men. Out of the 500 inhabitants that are not in the choir 300 are
men. What is the probability that a randomly drawn man is a member of the choir? Please
indicate the probability in percent.______ %
5a. Imagine we are throwing a five-sided die 50 times. On average, out of these 50 throws
how many times would this five-sided die show an odd number (1, 3 or 5)? ______ out of 50
throws.
5b. Imagine we are throwing a loaded die (6 sides). The probability that the die shows a 6 is
twice as high as the probability of each of the other numbers. On average, out of these 70
throws how many times would the die show the number 6? ________out of 70 throws.
6. In a forest 20% of mushrooms are red, 50% brown and 30% white. A red mushroom is
poisonous with a probability of 20%. A mushroom that is not red is poisonous with a
probability of 5%. What is the probability that a poisonous mushroom in the forest is red?
________
Cognitive reflection task (CRT)
1. A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does
the ball cost? _____cents
2. If it takes 5 machines 5 minutes to make 5 widgets, how long would it take 100 machines
to make 100 widgets? _____minutes
3. In a lake, there is a patch of lily pads. Every day, the patch doubles in size. If it takes 48
days for the patch to cover the entire lake, how long would it take for the patch to cover half
of the lake? ______days
25
Preliminary draft – please do not cite or distribute without authors permission
A2
Probit results
In order to make the results more interpretable and comparable to previous literature a
dummy variable was created out of the average of the financial management behavior scale
(FMBS). 1 indicating sound financial behavior corresponding to the mean value equal to or
above 4, and 0 otherwise. This allows for a probit analysis, with marginal effect for the probit
regressions are presented in table 10. In model (1) marginal effects indicate that, having one
more correct answer to the financial literacy questions is associated with a 6.5% increased
probability of engaging in desirable financial behavior. In model (5) self-reported financial
literacy as well as numeracy and CRT scores are included, here a one point increase on the
financial literacy quiz increases the probability of showing sound financial behavior by about
4.6%.
Table 10. Probit marginal effects
VARIABLES
Female
(1)
Marginal effects, at means
(2)
(3)
(4)
(5)
0.003
(0.020)
-0.058
-0.030
(0.020)
-0.053
-0.036*
(0.019)
-0.060
-0.000
(0.020)
-0.065
0.021
(0.019)
-0.089
(0.071)
0.003***
(0.073)
0.004***
(0.072)
0.004***
(0.071)
0.003***
(0.070)
0.002***
(0.001)
-0.004
(0.023)
(0.001)
-0.004
(0.023)
(0.001)
-0.005
(0.023)
(0.001)
-0.004
(0.023)
(0.001)
0.002
(0.023)
0.015
(0.025)
0.034
(0.025)
0.036
(0.025)
0.019
(0.025)
0.013
(0.025)
Income 25 000-34 999 SEK
0.007
(0.037)
0.127***
0.017
(0.038)
0.133***
0.020
(0.038)
0.136***
0.008
(0.037)
0.130***
-0.009
(0.038)
0.106***
Income 35 000-44 999 SEK
(0.036)
0.208***
(0.036)
0.226***
(0.036)
0.231***
(0.036)
0.210***
(0.036)
0.179***
(0.037)
0.147***
(0.042)
(0.037)
0.162***
(0.043)
(0.037)
0.165***
(0.043)
(0.037)
0.152***
(0.042)
(0.036)
0.119***
(0.042)
0.212***
(0.042)
0.224***
(0.042)
0.229***
(0.042)
0.216***
(0.042)
0.173***
(0.041)
Born in a non-European country
Age
Upper secondary education
University education
Income 15 000-24 999 SEK
Income 45 000-54 999 SEK
Income 55 000-64 999 SEK
26
Preliminary draft – please do not cite or distribute without authors permission
Income 65 000-74 999 SEK
Income>75 000 SEK
0.235***
(0.045)
0.208***
0.262***
(0.046)
0.225***
0.266***
(0.046)
0.230***
0.238***
(0.045)
0.210***
0.204***
(0.045)
0.170***
(0.042)
(0.042)
Financial literacy, number of correct answers
(0.042)
0.065***
(0.042)
0.071***
(0.042)
0.046***
(0.010)
-0.006
(0.007)
(0.010)
-0.006
(0.007)
-0.004
(0.011)
0.005
(0.011)
(0.009)
Numeracy score
0.013**
(0.005)
CRT score
0.016*
(0.009)
Self-reported financial knowledge
0.084***
(0.009)
Observations
2,062
2,062
2,062
2,062
2,062
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Determinants of financial behavior – an instrumental variable approach
Previous studies has considered the possibility of reversed causality with respect to financial
behavior and financial literacy. Knowledge about financial matters can explain ones
behavior, but it can also be that certain behavior leads to increased knowledge abot financial
matters. If an individual, for example, do not have much knowledge about the financial
markets but starts saving in stocks or mutual funds this persons financial knowledge can
clearly grow with this experience. A simple ordinary least square regression is not able to
account for such reversed causality, in order to minimize this influence it is possible to find
an instrument for financial literacy.
Numeracy, or at least math performance, has been used as an instrument for financial literacy
in previous work (Jappelli & Padula, 2013). But it might be argued that this trait is also
affected by a person’s financial behavior in the same manner as financial literacy. We
therefore also investigate the validity of cognitive reflection as an instrument. The results of
both the first stage regressions and the final 2SLS are presented in tables 11 and 12.
27
Preliminary draft – please do not cite or distribute without authors permission
Table 11. First stage regression, financial literacy score
OLS
VARIABLES
(1)
CRT score
Born in a non-European country
University education
Upper secondary education
Income 15 000-24 999 SEK
Income 25 000-34 999 SEK
Income 35 000-44 999 SEK
Income 45 000-54 999 SEK
Income 55 000-64 999 SEK
Income 65 000-74 999 SEK
Income>75 000 SEK
Observations
R-squared
(3)
0.286***
(0.0115)
(0.0433)
0.0871
(0.156)
0.0157***
(0.00130)
0.225***
(0.0554)
-0.0108
(0.0508)
0.0931
(0.0729)
0.0334
(0.0726)
0.226***
(0.0779)
0.134
(0.0889)
0.130
(0.0900)
0.365***
(0.0962)
0.203**
(0.0912)
0.141***
(0.0245)
0.241***
(0.0140)
(0.0430)
0.0819
(0.154)
0.0156***
(0.00129)
0.200***
(0.0551)
-0.0108
(0.0503)
0.0973
(0.0726)
0.0322
(0.0722)
0.227***
(0.0782)
0.112
(0.0883)
0.124
(0.0891)
0.357***
(0.0951)
0.205**
(0.0903)
2,062
0.378
2,062
0.387
0.382***
(0.0215)
Numeracy score
Age
(2)
(0.0453)
-0.0661
(0.154)
0.0140***
(0.00139)
0.289***
(0.0586)
-0.0137
(0.0537)
0.128*
(0.0767)
0.0875
(0.0767)
0.301***
(0.0861)
0.199**
(0.0937)
0.230**
(0.0941)
0.440***
(0.101)
0.301***
(0.0954)
2,062
0.299
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
28
Preliminary draft – please do not cite or distribute without authors permission
Table 12. 2SLS, Financial Management Behavior Scale (FMBS)
(1)
(2)
VARIABLES
Financial literacy, number of correct answers
Instrumented with CRT-score
0.080**
(0.035)
Financial literacy, number of correct answers
Instrumented with numeracy score
0.084***
(0.026)
Financial literacy, number of correct answers
Instrumented with CRT-score and numeracy score
Female
Born in a non-European country
Age
University education
Upper secondary education
Income 15 000-24 999 SEK
Income 25 000-34 999 SEK
Income 35 000-44 999 SEK
Income 45 000-54 999 SEK
Income 55 000-64 999 SEK
Income 65 000-74 999 SEK
Income>75 000 SEK
Observations
R-squared
(3)
0.083***
(0.026)
0.013
(0.035)
-0.034
(0.078)
0.006***
(0.001)
0.092**
(0.038)
0.009
(0.032)
0.107**
(0.047)
0.276***
(0.046)
0.391***
(0.051)
0.407***
(0.057)
0.505***
(0.057)
0.554***
(0.066)
0.523***
(0.058)
0.016
(0.031)
-0.034
(0.077)
0.006***
(0.001)
0.090**
(0.037)
0.009
(0.032)
0.106**
(0.047)
0.275***
(0.046)
0.390***
(0.050)
0.405***
(0.057)
0.503***
(0.056)
0.552***
(0.066)
0.521***
(0.057)
0.015
(0.031)
-0.034
(0.077)
0.006***
(0.001)
0.090**
(0.037)
0.009
(0.032)
0.106**
(0.047)
0.275***
(0.046)
0.390***
(0.050)
0.405***
(0.057)
0.504***
(0.056)
0.552***
(0.065)
0.522***
(0.057)
2,062
0.180
2,062
0.181
2,062
0.180
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
The Sanderson-Windmeijer (SW) first-stage F-statistics (equal to the Kleibergen-Paap Wald
statistic) are high for all specifications (316, 621 and 333 respectively), supporting non-weak
29
Preliminary draft – please do not cite or distribute without authors permission
instruments. In the third specification were the endogenous variable is overidentified the
results of the Sargan-Hansen test indicate that we cannot not reject the null, lending support
to the validity of the instruments. The results support previous findings that financial literacy
is significant, but the effect size is quite small.
30