SSRN Id4757297
SSRN Id4757297
d
The impact of better heuristics awareness to financial literacy.⋆
we
Francisco Pitthana,∗, Kristof De Wittea,b
a Leuven Economics of Education Research, Faculty of Economics and Business, KU Leuven
b UNU-Merit, Maastricht University
ie
Abstract
ev
A growing concern among policymakers is the enhancement of citizens’ financial well-being,
which necessitates improved financial decision-making and outcomes to alleviate individual fi-
nancial distress. While the prevailing approach to address this issue involves implementing
better financial education policies aimed at enhancing financial literacy, numerous other factors
r
influence financial decision-making, including behavioral biases, numeracy, and emotions. One
particular factor is the utilization of heuristics, or rule-of-thumb strategies, which serve as sim-
er
plifications of reality, facilitating decision-making in complex situations, particularly when time
is limited. However, the use of heuristics has been argued to result in suboptimal decisions when
uncertainty or long-term consequences are involved. This paper proposes a randomized con-
pe
trolled experiment to examine whether financial education programs that incorporate teachings
on heuristics, specifically availability heuristics and the gambler’s fallacy, can enhance students’
financial literacy. Our findings reveal that conventional financial education courses effectively
improved financial literacy to a significant extent, whereas courses that also covered heuristics
ot
did not yield similar results. This indicates that making the world more complex (i.e. by making
individuals realize their unconscious heuristics) can be hurtful to financial literacy, showing that
larger financial education curricula is not always better.
tn
⋆ This work was supported by the Research Foundation Flanders (FWO) (through fundamental research fel-
lowship under grant 11K4621N), and by the ‘Baloise Insurance Research chair to financial well-being’. This
Pr
experiment was registered in the American Economic Association’s registry for randomized controlled trials on
September 30, 2021 (AEA-0007956).
∗ Corresponding author.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
Do heuristics matter for financial literacy?
d
The impact of better heuristics awareness to financial literacy.
we
Abstract
ie
A growing concern among policymakers is the enhancement of citizens’ financial well-being,
which necessitates improved financial decision-making and outcomes to alleviate individual fi-
ev
nancial distress. While the prevailing approach to address this issue involves implementing
better financial education policies aimed at enhancing financial literacy, numerous other factors
influence financial decision-making, including behavioral biases, numeracy, and emotions. One
particular factor is the utilization of heuristics, or rule-of-thumb strategies, which serve as sim-
r
plifications of reality, facilitating decision-making in complex situations, particularly when time
er
is limited. However, the use of heuristics has been argued to result in suboptimal decisions when
uncertainty or long-term consequences are involved. This paper proposes a randomized con-
trolled experiment to examine whether financial education programs that incorporate teachings
pe
on heuristics, specifically availability heuristics and the gambler’s fallacy, can enhance students’
financial literacy. Our findings reveal that conventional financial education courses effectively
improved financial literacy to a significant extent, whereas courses that also covered heuristics
did not yield similar results. This indicates that making the world more complex (i.e. by making
individuals realize their unconscious heuristics) can be hurtful to financial literacy, showing that
ot
1. Introduction
rin
Insufficient financial literacy renders individuals susceptible to make poor financial choices, re-
sulting in adverse consequences, including an accumulation of excessive debt, bankruptcy, and
financial instability, perpetuating the cycle of poverty by the absence of knowledge and aptitude
required to effectively manage personal finances (Mitchell & Lusardi, 2015). In the absence of
ep
financial literacy, individuals may become susceptible to exploitative lending practices, high-
interest loans, and fraud, thereby trapping them in a recurring pattern of indebtedness and
restraining their financial autonomy (Khan et al., 2022). By lacking the proper financial knowl-
Pr
edge in a realm of complex and in-numerous financial instruments, consumers rely on heuristics
or rules of thumb, which are mental shortcuts that help them make fast financial decisions, but
can also result in errors (Bernheim, 1995; Lusardi & Mitchell, 2014). However, given the increas-
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
ing complexity of financial products and the amplified attention to provide costumers all relevant
d
information, it becomes fundamental to find effective methods to improve financial literacy and
to examine whether pointing individuals to heuristics changes their financial literacy (Lusardi,
we
2019; Shefrin & Nicols, 2014).
Financial education has emerged as the most effective approach thus far in enhancing financial
literacy. Extensive research indicates its positive impact on financial knowledge, and to a lesser
extent, on financial behavior, facilitating financial inclusion (Atkinson & Messy, 2013; Kaiser et
ie
al., 2022). However, the long-term effects of financial education on financial outcomes and overall
well-being remain uncertain. Existing findings are inconclusive and lack sufficient evidence to
ev
establish causal relationships (Kaiser et al., 2022; Willis, 2011), except for recent evidence from
Germany (Sutter et al., 2023). While alternative interventions like financial advisory services
have been proposed, their effectiveness is mixed, and empirical evidence is scarce (Chalmers &
Reuter, 2020). A major policy concern is related on which content should be included in financial
r
education curricula to improve the financial competences of the youth, especially considering
complex themes such as sustainable finance and investments (European Union/OECD, 2023).
er
Recent advancements in the literature highlight the significant role of behavioral and cognitive
factors in shaping financial literacy and decision-making processes (Jonsson et al., 2017; Pitthan
pe
& De Witte, 2021). Building upon this understanding, behavioral-based financial education
programs, i.e. when financial education also teaches how to make good financial decision and
about which factors could be in the way to make good decisions, show that learning about biases
can improve financial literacy (Pitthan & De Witte, 2023). Although often centred in behavioural
or cognitive biases, it is important to acknowledge that other factors, such as heuristics, can also
ot
lead to suboptimal outcomes. Heuristics serve as simplifications of reality that facilitate decision-
making in complex scenarios (Tversky & Kahneman, 1973). However, there is ambivalence
tn
al., 2014; Forbes et al., 2015; OECD, 2019), they can prove considerably detrimental in financial
decisions with long-term consequences, such as purchasing a house or choosing a pension plan
(Dhami et al., 2019; Quddus & Banerjee, 2021).
Given the increasing financial complexity, this paper examines the impact of financial edu-
ep
cation courses supplemented with heuristic materials on improving financial literacy. Moreover,
it explores if education can counteract the negative influence of heuristics. To test this, we
conducted a randomized controlled trial (RCT) with Belgian secondary school students. We
Pr
compared traditional financial education materials to modified versions including content on the
availability heuristic and the gambler’s fallacy (Kahneman et al., 1971; Tversky & Kahneman,
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
1973). We assess whether the effect of behavioral-based financial education on financial literacy
d
can be mediated by increased awareness of heuristic behavior. Contrary to assumptions, only
traditional financial education courses improved financial literacy, while behavioral courses with
we
heuristic content did not. We did not observe a similar indirect effect to financial literacy medi-
ated by heuristics, as seen with the myopic bias (Pitthan & De Witte, 2023). These findings can
be attributed to the dual nature of heuristics, which can both simplify reality and undermine
decision-making in complex scenarios. Consequently, teaching about heuristics may not be the
ie
optimal delivery method, as it can potentially confuse students by emphasizing both the use-
fulness and the detrimental effects of heuristics on decision-making skills. This makes decisions
ev
more complex, since the simplifications from heuristics now are no longer unconscious, which
worsens financial literacy outcomes.
r
The literature presents causal frameworks that propose connections among financial education,
er
financial literacy, and financial outcomes (Carpena & Zia, 2020; Pitthan & De Witte, 2023).
Since heuristics can play a role on financial decisions (Strough et al., 2011), we hypothesize
that such causal frameworks could be augmented in order to also consider the role of heuristics
pe
in financial literacy and financial decisions. We propose the following mechanism: behavioural-
based financial education courses that teach about heuristics would make individuals more aware
of such rules of thumb. In turn, this could make them more conscious of their financial decisions,
improving their financial literacy and financial outcomes. Our hypothesized framework is shown
ot
Figure 1. Direct acyclic graphs with estimated causal links of behavioural-based financial education mechanisms.
rin
The experiment is centred in financial education courses offered to secondary schools in the
Dutch-speaking part of Belgium, with students aged between 14 and 19 years.1 The courses are
in the format of computer-assisted games, covering the main financial literacy objectives from
the Flemish curriculum, with articles, games, videos, and exercises. Students make financial
ep
decisions in close to real life examples, receiving feedback on their mistakes. The experiment
has three different experimental arms (to avoid contamination randomized at the school level).
(i) A control group that does not receive any course until the end of the experiment. (ii) A
Pr
treatment group that receives traditional financial education (TFE), i.e. a version of the game
1 The experiment was conducted between September 2021 and January 2022.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
focusing only on financial knowledge concepts. (iii) A treatment group that receives heuristics
d
financial education (HFE), i.e. a modified version of the game, including behavioural-based
material about the heuristics of availability and gambler’s fallacy, in addition to some financial
we
knowledge content. Students in one experimental arm only follow the course from their respective
treatment (if any). All participants fill out both pre and post-tests, which include socioeconomic
questions; financial literacy tests with questions inspired on OECD/INFE (2011); and heuristics
tests, partially based on Camerer and Kunreuther (1989). An overview of a few examples seen
ie
in the course is given in Appendix A.
We estimate the average treatment effects (ATE) and average direct effect (ADE) of our
ev
treatment variables (i.e. receiving either the TFE or HFE courses) to our main outcome variable
(i.e. financial literacy) and to our mediating variable (i.e. better heuristics awareness) using OLS
estimates, clustering standard errors by school. The treatments are constructed as two dummy
variables, with a value of 1 assigned to indicate if participated in a certain treatment, while
r
utilizing the control group as the baseline when both dummy variables are set to 0. The causal
mediation analysis follows the approach of Imai et al. (2010), in order to estimate average causal
er
mediated effects (ACME).2 The ATE and ADE are identified by the statistical independence
of the treatment status with both the mediation and outcome variables, which is given by the
pe
random attribution of treatments from our RCT design. The statistical independence of the
mediation and outcome variables (that identifies the ACME) is indirectly tested by sensitivity
analysis and proxy tests for endogeneity.
3. Results
ot
Table 1 shows the ATEs for the main variables of our study, financial literacy and heuristics
scores. Although the treatment achieved full compliance following exogenous allocation, we
tn
still observe sample imbalance due to attrition in the post-tests. To address this issue, we use
both Mahalanobis distance matching (MDM) and Lee (2009)’s bounds techniques, in addition
to controlling for any balancing differences.3 In Panel A, we observe that all ATEs to financial
literacy are positive, although only the traditional financial education has a significant effect of
rin
lation to the main financial literacy outcomes. Alternatively, it is plausible that the explanation
2 For similar causal mediation applications in the financial education literature, see Carpena and Zia (2020)
and Pitthan and De Witte (2023).
Pr
3 A post-hoc power analysis found a minimum detectable effect-size of 0.31, considering a power of 80%, sample
of 427, significance level of 5%, 42 clusters and the estimated intraclass correlation of clusters of 0.01. An overview
about the data and sample is seen in Appendix B.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
of heuristics could cause confusion among students as they become suddenly aware of the com-
d
plexity involved in making decisions. Besides, we note in Appendix B that the financial literacy
pre-test score of students that participated in the heuristics education treatment is significantly
we
higher in comparison to the other treatment arms. Although we control for this difference in
the analysis, we do not exclude the hypothesis that this could undermine the total effect of the
course.
From Panel B, we observe that students in the traditional financial education treatment
ie
experienced a significant increase in their heuristics behaviour by 0.93 sd when compared to the
control group.4 The observed effect size is notably substantial and may be associated with the
ev
hypothesis that improving financial literacy in the absence of exposure to heuristics could lead
students to develop increased confidence in their decision-making, thereby exhibiting a greater
inclination towards relying on heuristics. In the case of heuristics education as the treatment,
the ATE remains positive, although it is not statistically significant.
r
Table 1. Estimated ATEs and Lee bounds of main variables.
Specification
er Mahalanobis
Matching
Lower Bound Upper Bound Trimming proportion
F 147.1
Note: Lower and upper bounds coefficients are the Lee (2009)’s bounds estimations for the required trimming rate in comparison with control conditions.
Treatments are constructed as dummy variables, indicating 1 if followed a certain treatment, and 0 otherwise. Numbers in brackets are the 95% confidence
intervals. The Mahalanobis Distance Matching estimated coefficients are in terms of standard deviations of the outcome variable. Clustered standard errors in
parentheses. * p ≤ .05 ** p ≤ .01 *** p ≤ .001. Estimation controls for all pre-test observed characteristics presented at Appendix B.
In Table 2, we directly examine the hypothesis presented in Figure 1, which explores whether
rin
education that enhances awareness of heuristics can lead to an indirect improvement in financial
literacy. Our findings indicate that we have evidence to reject this hypothesis. Neither of the two
treatments exhibited a significant average causal mediation effect (ACME) in enhancing financial
ep
literacy scores, although the ACME is positive for the heuristics education treatment. Notably,
the ACME for the traditional financial education case is even negative, although it is not statis-
tically significant. This observation may be associated with the notable rise in heuristics-based
behavior observed in Table 1. Furthermore, both the average direct effect (ADE) and the aver-
Pr
4 From Appendix C, this seems to be driven mostly by the availability heuristics type of behaviour.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
age treatment effect (ATE) are only statistically significant for the traditional financial education
d
case, aligning with the findings presented in Table 1. To provide further support, Appendix C
replicates the results of both tables for each of the two individual heuristics (availability and
we
gambler’s fallacy). Additionally, Appendix D discusses the sensitivity analysis and endogeneity
tests conducted to assess the robustness of our mediation analysis.
Table 2. Mediation analysis for financial literacy post-test as outcome with effect mediated by heuristics.
ie
Treatment Trad. Treatment
Fin. Educ. Heuristics Educ.
-0.009 0.008
Average Causal Mediated Effect (ACME)
(0.020) (0.013)
ev
0.476∗∗∗ 0.132
Average Direct Effect (ADE)
(0.150) (0.116)
0.466∗∗∗ 0.140
Average Treatment Effect (ATE)
(0.136) (0.116)
% of ATE Mediated −2.00% 3.92%
N 249 283
r
R2 0.356 0.300
F 166.21 21.86
Note: The results show direct and indirect effects estimated by mediation analysis using the algorithm proposed by
Imai et al. (2010). Standard errors obtained by quasi-Bayesian Monte Carlo simulations. * p ≤ .05 ** p ≤ .01 ***
er
p ≤ .001. Estimation controls for all pre-test observed characteristics presented at Table 2.
4. Discussion
pe
In this paper, we examined the potential role of heuristics in enhancing financial literacy. To
investigate this, we conducted a randomized controlled trial in Belgium involving 427 secondary
school students. The results revealed that students who participated in traditional financial edu-
cation courses experienced a significant improvement in their financial literacy scores. However,
ot
this improvement was not observed among students who received instruction on heuristics. Ad-
ditionally, our mediation analysis indicated that increased awareness of heuristics does not serve
tn
hand, heuristics can be beneficial by simplifying complex situations. On the other hand, they
can be detrimental by making individuals more prone to making mistakes. In essence, teaching
about heuristics can be too complicated for students, leaving them uncertain about the benefits
ep
and drawbacks, while potentially diverting their attention away from the main course content.
The results presented in Panel B of Table 1 indicate that traditional financial education may
contribute to an increase in heuristics-based behavior among students. This may be attributed
to the confidence gained from the course. However, despite this shift in behavior, it does not
Pr
appear to have a negative impact on students’ financial literacy, as evidenced by the absence of
any mediation effect of heuristics, as shown in Table 2.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
One lingering question remains: why does the improved awareness of heuristics not have a
d
statistically significant indirect effect on enhancing financial literacy, unlike the positive impact
observed with the learning of behavioral biases such as myopia (Pitthan & De Witte, 2023)?
we
First, heuristics should not be viewed solely as a negative influence on financial decisions. This
is especially evident when individuals are faced with time constraints, as heuristics can offer a
practical solution. Second, some behavioural biases like myopia can have a more direct link to low
financial literacy (e.g. short-term preference), which is not always seen in the case of heuristics.
ie
Finally, the realization of exactly when heuristics are useful or dangerous might be more beneficial
to individuals, instead of better awareness to different types of heuristic behaviour.
ev
In terms of policy implications, our findings suggest that the inclusion of heuristics in financial
education programs may not be advisable, as it appears to diminish the effectiveness of such
courses, making decisions more complex than they should be. Although recent national strategies
(US Treasury, 2020) are built given a very complex financial system, the answer to effective
r
education programs is not to make decisions more difficult than they should be. Furthermore,
considering the dual nature of heuristics, policies aimed at mitigating their negative effects should
er
prioritize interactive approaches, such as ”learning by doing,” rather than traditional teaching
methods. By engaging in practical exercises and experiencing firsthand both the advantages and
pe
disadvantages of heuristics, individuals can develop a more comprehensive understanding of their
implications.
Our analysis has some limitations. First, the limited sample size resulted in a power that was
small. Second, although we tested our hypothesis with both the availability and gambler’s fallacy
heuristics, many other heuristics were not analysed (e.g. framing effect, affection, representa-
ot
tiveness), which might have a similar or different effect, depending on how closely linked with
financial literacy they are. Third, the analysis was realized in a time span of 4 months, which
tn
does not allow to draw conclusions about long lasting outcomes of financial education courses.
Finally, the hypothesis of causal mediation analysis are not directly testable, although they were
still indirectly tested here with proxy measures and sensitivity analysis. Future research should
address these points, by conducting experiments in different locations, with different heuristics
rin
and biases, and especially, with a longer observational period, in order to measure long lasting
financial outcomes and well-being.
References
ep
Atkinson, A., & Messy, F. A. (2013). Promoting financial inclusion through financial education.
OECD/INFE evidence, policies and practice, 34.
Bernheim, D. (1995). Do households appreciate their financial vulnerabilities? an analysis of
Pr
actions, perceptions, and public policy. Tax policy and economic growth, 3, 11–13.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
Camerer, C. F., & Kunreuther, H. (1989). Decision processes for low probability events: Policy
d
implications. Journal of policy analysis and management, 8 (4), 565–592.
Carpena, F., & Zia, B. (2020). The causal mechanism of financial education: Evidence from
we
mediation analysis. Journal of Economic Behavior & Organization, 177, 143–184.
Chalmers, J., & Reuter, J. (2020). Is conflicted investment advice better than no advice? Journal
of Financial Economics, 138 (2), 366–387.
Dhami, S., Al-Nowaihi, A., & Sunstein, C. R. (2019). Heuristics and public policy: Decision-
ie
making under bounded rationality. Studies in Microeconomics, 7 (1), 7–58.
Drexler, A., Fischer, G., & Schoar, A. (2014). Keeping it simple: Financial literacy and rules of
ev
thumb. American Economic Journal: Applied Economics, 6 (2), 1–31.
European Union/OECD, C. (2023). Financial competence framework for children and youth in
the european union. EU Commission & OECD Publishing.
Forbes, W., Hudson, R., Skerratt, L., & Soufian, M. (2015). Which heuristics can aid financial-
r
decision-making? International review of Financial analysis, 42, 199–210.
Gigerenzer, G., & Gaissmaier, W. (2011). Heuristic decision making. Annual review of psychology,
62, 451–482.
er
Imai, K., Keele, L., & Tingley, D. (2010). A general approach to causal mediation analysis.
pe
Psychological methods, 15 (4), 309.
Jonsson, S., Söderberg, I.-L., & Wilhelmsson, M. (2017). An investigation of the impact of fi-
nancial literacy, risk attitude, and saving motives on the attenuation of mutual fund
investors’ disposition bias. Managerial Finance, 43 (3), 282–298.
Kahneman, D., et al. (1971). Belief in the law of small numbers. Psychological bulletin, 76 (2).
ot
Kaiser, T., Lusardi, A., Menkhoff, L., & Urban, C. (2022). Financial education affects financial
knowledge and downstream behaviors. Journal of Financial Economics, 145 (2).
tn
Khan, F., Siddiqui, M. A., & Imtiaz, S. (2022). Role of financial literacy in achieving financial
inclusion: A review, synthesis and research agenda. Cogent Business & Management,
9 (1), 2034236.
Lee, D. S. (2009). Training, wages, and sample selection: Estimating sharp bounds on treatment
rin
OECD. (2019). Tools and ethics for applied behavioural insights: The basic toolkit.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297
OECD/INFE. (2011). Measuring financial literacy: Core questionnaire in measuring financial
d
literacy: Questionnaire and guidance notes for conducting an internationally comparable
survey of financial literacy.
we
Pitthan, F., & De Witte, K. (2023). How learning about behavioural biases improves financial
literacy. Available at SSRN, 4360086.
Pitthan, F., & De Witte, K. (2021). Puzzles of insurance demand and its biases: A survey on
the role of behavioural biases and financial literacy on insurance demand. Journal of
ie
Behavioral and Experimental Finance, 30, 100471.
Quddus, K., & Banerjee, A. (2021). Understanding heuristics-based financial decision-making
ev
using behavioral portfolio strategies. Review of Behavioral Finance.
Shefrin, H., & Nicols, C. M. (2014). Credit card behavior, financial styles, and heuristics. Journal
of Business Research, 67 (8), 1679–1687.
Strough, J., Karns, T. E., & Schlosnagle, L. (2011). Decision-making heuristics and biases across
r
the life span. Annals of the New York Academy of Sciences, 1235 (1), 57–74.
Sutter, M., Weyland, M., Untertrifaller, A., Froitzheim, M., & Schneider, S. O. (2023). Financial
er
literacy, experimental preference measures and field behavior–a randomized educational
intervention. IZA Working Paper Series.
pe
Tversky, A. (1996). On the reality of cognitive illusions. Psychological Review, 103 (3), 582–591.
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and prob-
ability. Cognitive psychology, 5 (2), 207–232.
US Treasury, D. o. (2020). U.s. national strategy for financial literacy. US Financial Literacy &
Education Commission, Washington, DC.
ot
Willis, L. E. (2011). The financial education fallacy. American Economic Review, 101 (3).
tn
rin
ep
Pr
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4757297