Financial Knowledge and Finance Behaviors in Latin America and its Relationship with the COVID-19 Pandemic Period

doi:10.21203/rs.3.rs-2355441/v2

Download PDF

Research Article

Financial Knowledge and Finance Behaviors in Latin America and its Relationship with the COVID-19 Pandemic Period

https://doi.org/10.21203/rs.3.rs-2355441/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted

You are reading this older preprint version

Read the latest preprint version →

Governments worldwide conceive financial education as a crucial tool for enhancing the financial well-being of families. However, the relationship between financial knowledge and financial behaviors remains inconclusive, particularly in developing countries where a limited number of studies have explored this connection. This study employed a subjective and an objective measure of financial knowledge and assessed 16 financial behaviors of participants from Argentina, Colombia, Ecuador, Mexico, and Peru. Out of these 16 financial behaviors, 14 have a relationship with at least one of the two measures. This study found positive correlations between financial knowledge and various financial behaviors, including planning behaviors, acquisition of formal financial products, willingness to assume risks, seeking financial advice, and cautious management of their financial products. The relationship between the acquisition of informal financial products was positive with the objective financial knowledge measure but negative with the subjective measure. The identification of causal relationships proved challenging and often yield inconclusive results due to identified endogeneity issues and the presence of weak instruments. In general, during the COVID-19 period, participant displayed increased prudency in their financial practices and managing their products. This effect was larger among participants with higher levels of financial knowledge. Future research in Latin America should incorporate the distinction between financial literacy and financial knowledge, explore qualitative research designs, and evaluate the effectiveness of financial educational programs.

JEL classification codes: G51, G53

Behavioral Economics

Financial Knowledge

Financial Literacy

Household Finance

Consumer Behavior

Financial Education

Behavioral Finance

Huston (2010) defines financial knowledge dimension as the “stock of knowledge acquired through education and/or experience specifically related to essential personal finance concepts and products”. In contrast, Vieira et al. (2019) define financial literacy as a set of beliefs, attitudes, and behaviors that allow and enable people to make responsible financial decisions. Although, these concepts are distinct, in practice, many studies have blurred the distinction between them when measuring them (Huston, 2010). Consequently, most studied cited in this document present their results in terms of financial literacy, but the term is changed to “financial knowledge” when the measurement emphasizes on the participant’s knowledge dimension.

A substantial body of knowledge highlights the negative relationship between the lack of financial knowledge and the achievement of personal finance goals such as wealth accumulation and retirement (Gomes et al., 2021; van Rooij et al., 2012). Given the importance of financial knowledge in personal well-being, governmental institutions around the world have designed strategies to promote financial education among vulnerable populations, including single mothers, young people, and low-income families, in order to improve their financial decision-making and foster their inclusion in the financial system (Lusardi & Mitchell, 2007, 2014; Mitchell & Lusardi, 2015). Furthermore, the connection between financial education and financial behaviors has been extended to the concept of financial well-being (Brüggen et al., 2017; Gutter & Copur, 2011).

Given the relevance of this topic, most governmental programs aim to use financial education for two primary purposes (Atkinson & Messy, 2013): 1) to increase financial inclusion, examples of this can be found in countries like Mexico, Peru, Brazil, the UK, Ireland, or Colombia, and 2) To enhance financial consumer protection, as seen in countries like Indonesia, Malaysia, or India. These programs gained prominence during the COVID-19 pandemic. For instance, the US National Strategy for Financial Literacy 2020 became an institutional source of information providing guidance to US citizens facing financial distress during the COVID-19 pandemic (Literacy & Commission, 2020).

As a consequence, academic studies in financial literacy are making progress in understanding how financial education programs must be structured in order to have a significant impact on people’s financial behavior. For instance, Fernandes et al. (2014) found that financial education interventions by third parties explains only 0.1% of financial behaviors. In addition, financial education programs should also consider the role of social context and behavioral traits in order to evaluate the capacity of any family to make sound financial decisions and how these capacities might be improved (Badarinza et al., 2016; Guiso & Sodini, 2013; Santini et al., 2019). For example, financial education interventions targeted to vulnerable populations often have a limited impact on savings habits, as these population may lack the resources to save (Frisancho, 2016).

This study aims to address a current need of research on the relationship between financial knowledge and financial behaviors in Latin America, as proposed by Goyal & Kumar (2021). The authors acknowledge the scarcity of publications addressing the relationship between different measures of financial literacy, including its knowledge dimension, and various financial and economic characteristics of people in Latin American countries, specifically Argentina, Colombia, Ecuador, Mexico, and Peru. Two measures of financial literacy are employed: one objective measure, using “The Big Three” questions (interest rate, inflation, and diversification), and subjective measure on a scale from one to five (Gomes et al., 2021).

The results indicate that out of 16 financial behaviors, 14 showed the expected relationship with either of the two measures. There is a positive correlation between financial knowledge and planning behaviors, acquisition of formal financial products, willingness to assume risks, seeking financial advice, and cautious management of their financial products. The acquisition of informal financial products shows a positive association with the objective measure of financial knowledge but a negative correlation with the subjective measure.

Additionally, estimations did not find sufficient evidence to stablish a relationship between either of the financial knowledge measures and having a savings account or paying bills on time. The COVID-19 pandemic played a more significant role in the relationship between the objective financial knowledge measure and financial behaviors than the subjective measure. In general, participants were more cautious with their financial behaviors during the pandemic, with this effect being larger among those categorized as financially knowledgeable. Drawing definitive causal relationships from the results is challenging due to the lack of strong instruments in most cases.

Although the results support the idea that financial education is crucial for promoting better financial practices, we advocate for future research with qualitative component and the evaluation of financial education programs. Furthermore, it is important to include the distinction between financial literacy and financial knowledge, especially in future studies aiming to understand how educational programs impact participants’ behaviors.

The present study is organized as follows: Section 2 provides a brief literature review on the impact of financial knowledge on financial behaviors; Section 3 presents the data; Section 4 outlines the methodology; Section 5 presents the results; Section 6 offers a discussion, and Section 7 presents the conclusions.

Several meta-analyses have been conducted on financial knowledge worldwide. For instance, Hastings et al. (2013) examined whether financial education improves personal financial outcomes, primarily based on US data. Their overall finding suggests that the empirical relationship between the two factors is inconclusive. Similarly, Fernandes et al. (2014) conducted a meta-analysis and found a positive bias in the effect of financial literacy on financial behaviors. However, they observed that this effect is weaker or absent among individuals with low income and when using instrumental variable methodology. While Fernandes et al. (2014) did not report country-specific information, it can be found only two studies from Mexico regarding Latin America in their references.

On the other hand, other meta-analyses have discovered a strong relationship between financial knowledge and certain financial behaviors. Miller et al. (2015) utilized international data and found that financial education programs influence financial behaviors such as regular savings and record-keeping but did not have a significant impact on reducing the probability of loan default. However, when the data were segmented by country, specifically for Latin America, they found that these educational interventions positively impacted saving behaviors in Brazil, while there was no evidence of financial education affecting saving habits or reducing the likelihood of loan default.

Santini et al. (2019) also conducted a meta-analysis and found a positive relationship between financial knowledge, which has a positive impact on the person’s financial literacy, and behaviors such as seeking lower financial costs in loans and displaying lower risk aversion. In their sample, they only used data from Brazil, representing a single country in Latin America. The study acknowledges that the varying relationships between financial literacy and financial behaviors may be explained by regional segmentation. The results are influenced by the structure of each financial education program and cultural factors, emphasizing the importance of qualitative evidence to fully understand these relationships.

Another comprehensive study conducted by Kaiser & Menkhoff (2020) analyzed 37 experimental studies in schools, three of which were from Latin America, and also concluded that financial education has an impact on financial behaviors. Consistent with these results, Kaiser & Menkhoff (2017) found similar outcomes, although not for all financial behaviors, particularly those related to debt management. Out of 126 studies, only three were from Latin America.

Despite the fact that most of the previous studies do not cover comprehensively countries from Latin America, several non-meta-analysis studies have registered low levels of financial knowledge and financial literacy in Latin America. Moreover, the evidence regarding its impact on financial outcomes is scarce and tends to fail to establish a clear relationship between the two variables (Frisancho, 2016; Karakurum-Ozdemir et al., 2019).

Furthermore, Prado et al. (2022) conducted a systematic review of financial literacy studies in countries from Latin America and The Caribbean (LAC) from January 2016 to January 2022. They found that from 3,008 publications worldwide on financial literacy, only 65 were focused on LAC countries, with 41% being published in 2021. The most studied country was Brazil, ignoring endogeneity issues in most cases. For other countries in Latin America, the number of studies covered in Prado et al. (2022) were as follows: 15 studies from Mexico, nine from Chile, eight from Colombia, four from Ecuador, two from Peru, and none from Argentina.

In the case of Colombia, most of the studies on financial knowledge surveyed students (Prado et al., 2022) and they ignored the issue of endogeneity when establishing a relationship between financial knowledge and several financial behaviors. For instance, Cao-Alvira et al. (2021) found a statistically significant relationship between financial knowledge and behavior variables such as debt utilization, having a mortgage, debt-to-income ratio, and net worth. Another study by Muñoz-Murillo et al. (2020) found that the degree of the student’s cognitive abilities significantly impacts financial knowledge, and this relationship remain significant even after controlling for common confounding factors.

In the case of Mexico, Hernández-Mejía et al. (2021) established a positive relationship between owning a credit card and financial knowledge. García Mata (2021) states that a higher level of financial literacy increases the intention to engage in active financial planning for retirement purposes.

Chile has some important publications on financial knowledge, but this variable was not the primary subject of research (Prado et al., 2022). There are other studies that assess the impact of education programs on the levels of financial literacy among students (Martínez A & Puentes, 2018) and the relationship of financial knowledge with behaviors related to savings, paying bills on time, and budgeting (de la Fuente-Mella et al., 2021).

For Perú and Ecuador, the number of studies attempting to assess the relationship between financial knowledge and financial behaviors is scarce, but some related studies can be found, such as the ones by Bernal & Olivera (2020) and Prado et al. (2019).

Finally, the literature review by Goyal & Kumar (2021) identifies some emerging themes in financial literacy, such as how to improve financial literacy indicators in regionally dependent contexts, understanding the expected or intended outcomes for financial literacy and education programs in developed countries, and addressing the source of the financial literacy gender gap. The present study aims to provide evidence about the relationship between financial knowledge and financial behaviors in selected Latin American countries. The distinction about financial literacy and financial knowledge is addressed in section 4.

The study collected 2,213 observations from an online survey linked to a free web course about personal finance on the edX platform[1]. Participation was voluntary, and participants could obtain a certification for a fee. The surveys were collected between January 2019 and Abril 2021, considering responses from one year before and after the onset of the COVID-19 spread in Latin America. We only considered surveys with complete information concerning our independent and dependent variables.

To avoid response bias, we excluded surveys that took less than eight minutes and longer than 24 minutes to complete. Additionally, participants younger than 20 years old were excluded, as their involvement in certain financial behaviors would be limited (Huston, 2010). Furthermore, our study was limited to the countries with the largest number of surveys completed: Argentina (176), Colombia (1,401), Ecuador (127), Mexico (330), and Peru (179).

In March of 2020, most of the analyzed countries implemented their initial measures to control the spread of the COVID-19. Peru and Ecuador imposed significant mobility restrictions on March 16, followed by Argentina and Colombia on March 20 and 25, respectively. Mexico did not implement strict mobility restrictions; however, on March 23, they stablished a public program to control the spread of the COVID-19 virus by suspending public events and implementing social distancing programs. Consequently, a COVID-19 indicator was defined as one if the survey was answered before March 2020 and zero if it was answered after March 2020. Observations from March were excluded.

From the selected sample, the average respondent age was 31 years, 55% of the respondents being male, and 36% married. The majority of the sample came from urban areas (94%), and they had a stable income (78%). On average, a household had 0.6 children and lived with 1.8 additional adults. Regarding education, 53% of the respondents had an undergraduate degree as their highest education level, and 55% of the observations had a monthly income above the sample country’s average. Table 1 presents the descriptive statistics of the sample, along with the definitions of the variables[2].

Table 1. Descriptive statistics.

Panel A	Mean	Std. Dev.	Min	Max
COVID	0.80	0.40	0.00	1.00
Age	31.43	9.13	20.00	74.00
Male	0.55	0.50	0.00	1.00
Married	0.36	0.48	0.00	1.00
# Children live with	0.57	0.87	0.00	8.00
# People over 18 live with	1.77	1.37	0.00	10.00
Income above average	0.55	0.50	0.00	1.00
Stable Income	0.78	0.42	0.00	1.00
Urban	0.94	0.23	0.00	1.00
Panel B	Percentage (%)	Cumulative (%)
Education:
None	0.99	0.99
High School	33.39	34.39
Undergraduate Degree	53.14	87.53
Master's Degree	12.07	99.59
Doctorate Degree	0.41	100

Note: “COVID” is a dummy variable that takes the value of 1 if the survey was filled out after March 2020 or 0 if it was filled out before March 2020. “Male” is a dummy variable that takes the value of 1 if the person is male or 0 if the person is female. “Married” is a dummy variable that takes the value of 1 if the person lives with a partner or 0 if the person is single, those who are including divorced or widowed. “# Children live with” indicates the number of children (under 18) who live with the person. “# People over 18 live with” indicates the number of people over 18 who live with the person. “Income above average” is 1 if the income range is above the average range of the country where the person lives, and 0 otherwise. “Stable income” is 1 if the person receives labor benefits such as paid vacations or social security, and 0 otherwise. “Urban” is 1 if the person lives in an urban area, and 0 otherwise. The total number of observations for the study is 2,213.

The sample is relatively large; as noted by Huston (2010), the average sample size in studies related to financial literacy is 1,575 observations from 72 studies. Despite this advantage, the sample is subject to self-selection bias, limiting the generalizability of the conclusions. However, some inferences can be drawn for a broader population. Section 4.3 provides a more detailed explanation of the estimation methodology and its limitations.

When comparing the statistics in Table 1 with the data from Comisión Económica para América Latina y el Caribe (CEPAL), in 2022, CEPAL reported an average age of 38.7, and the percentage of the male population was 49.0% for individuals aged between 15 to 74 in Argentina, Colombia, Ecuador, Mexico, and Peru. In terms of the education level, the CEPAL reported that from 2020 to 2022, 72.9% of the population had 12 years of education or lower in Colombia, Ecuador, Mexico, and Peru. Therefore, participants tend to be younger, male, and have a higher level of education compared to the general population.

In section 4.1, two financial knowledge measures employed in this study are presented; in section 4.2, a list of financial behaviors considered for analysis is elaborated; and in section 4.3, the estimation methods are explained.

4.1 Measuring financial knowledge

The literature on financial knowledge uses three general questions to measure respondent’s financial knowledge, commonly known as “The Big Three” (Behrman et al., 2012a; Gomes et al., 2021; Lusardi & Mitchell, 2014; Mitchell & Lusardi, 2015). The objective financial knowledge measure takes the value of one if all three questions are answered correctly; otherwise, the value is zero.

The three general questions with their correct answer are[1]:

Interest rate: Suppose you deposit $100,000 in a saving account with an interest rate of 2% per year. You do not make any withdrawals or payments with that money. How much would you have at the end of the fifth year is: (a) more than $110,000 (correct choice); (b) exactly $110,000; (c) less than $110,000; (d) it is impossible to know with the given information; (e) don’t know/don’t answer.
Inflation: Imagine that someone must wait one year to get $1’000,000, and the country’s inflation rate is 2% per annum. After a year, this person can buy: (a) More than what he can buy now; (b) the same quantity than what he can buy now; (c) less than what he can buy now (correct choice); (d) depending on what he is going to buy; (e) don’t know/don’t answer.
Diversification: It is less probable that you lose all your money if it is invested in more than one place: (a) false; (b) true (correct choice); (c) don’t know/don’t answer.

Additionally, subjective financial knowledge is measured thorough self-reporting as follows: “I have good knowledge of finance in general terms”. The answer is on a scale of one to five, where one means that the participant disagrees and five indicates strong agreement with the sentence.

Nicolini & Cude (2022) describe some challenges in quantifying a person’s financial literacy and discuss the difference between financial literacy, financial knowledge, and financial capability. According to Huston (2010), financial literacy refers to how well an individual can understand and utilize financial information, while financial knowledge represents the understanding aspect of financial literacy. Therefore, the two measures employed in this study are more closely aligned with financial knowledge. In the literature review of section 2, several cited studies do not consider the difference and used the two terms interchangeably.

The discussion surrounding the proper measurement of financial literacy, which integrates both understand and the application of financial knowledge, is still in a developmental stage (Dew & Xiao, 2011; Knoll & Houts, 2012; Ranyard et al., 2020; Remund, 2010; Schmeiser & Seligman, 2013). The financial knowledge objective and subjective measures applied in this study were more motivated by their wide application in other studies and organizations. Also, they were selected to both keep concise and reduce survey fatigue, while also ensuring that the results are related to the financial knowledge dimension of financial literacy.

4.2 Financial behaviors

The survey asked participants about the financial products they had, how they managed them, and how they handled their income. Table 2[2], presents eight groups of financial behaviors with their expected relation to financial knowledge. The term “financial behaviors” is preferred over “financial practices” because it encompasses concepts such as willingness and planning, not just what the participants possess.

For planning behaviors, the list includes whether participants pay their bills on time, compare competing financial products before acquiring them, have a budget, and consider affordability before making purchases.

Regarding the acquisition of formal financial products, the list includes mortgages, stock market investments, credit cards, and savings accounts. For informal financial products, the survey considered informal loans and “cadenas”, which involve a group of people who periodically contribute the same amount of money to one group member in turns, without recognizing interests.

Other categories inquired about participant’s willingness to assume risks, whether they had sought financial assistance, set long-term goals, had a negative income balance, and how they managed their financial products. Specifically, if they had used an overdraft in the last year and if they regularly set money aside in their savings accounts.

Table 2. Financial behaviors to study.

Financial behaviors groups	List	Authors
Planning behaviors	Consider before buy (+) Pay bills on time (+) Compare before buy (+) Have budgets (+)	Kim et al. (2019); Lusardi & Mitchell (2007, 2017); van Rooij et al. (2011)
Acquisition of financial products	Have a mortgage (+) Have stock market investments (+) Have credit card (+) Have savings account (+)	Jappelli & Padula (2013); Kim et al. (2019); Klapper et al. (2013); Lin et al. (2017); van Rooij et al. (2011)
Willingness to assume risks	Willingness to take risks (+)	Campbell (2006); Muñoz-Murillo et al. (2020); M. van Rooij et al. (2011)
Informal sources of borrowing	Have informal loans (-) Save in “cadenas” (-)	Klapper et al. (2013); Santos et al. (2018)
Search for financial advice	Financial assistance (+)	Collins (2012); von Gaudecker (2015)
Management of financial products	Set aside money to save (+) Financial overdraft (-)	Gerardi et al. (2013); Guiso & Sodini (2013); Kim et al. (2019); Lusardi & Tufano (2015)
Setting long-term financial goals	Set long-term goals (+)	Kim et al. (2019)
Spend higher than income	Spends higher than income (-)	Kim et al. (2019); Klapper et al. (2013)

Note: The table lists the independent variables used in the estimations. A “+” sign indicates that an expected positive relationship between the financial behavior and financial knowledge, while a “-” sign indicates an expected negative relationship.

4.3 Estimations methods and limitations

Measurement error can be a concern because financial knowledge cannot be directly measured. Therefore, this study work with two measures. Additionally, there may be potential selection bias since as our sample consist of people voluntarily and actively participating in an online course about personal finance. Nevertheless, the results can still serve as a relevant reference for future financial education programs in Latin America and help understand the relationship with various financial behaviors, despite considering the limitations of the results.

The regression model is presented in the equation (1), and it is estimated using a Logit model for binary dependent variables or Ordered Logit model for categorical dependent variables.

[1] Including the option “Don’t know” is essential to reduce random answers and obtain a better measure of financial knowledge.

[2] The formulated questions in Spanish and English are presented in Appendix 2.

In (1), is the financial behavior being analyze, and the complete list is provided in Table 2, is the financial knowledge measurement, with two options considered: the objective and the subjective measurement. represents a list of control variables, including age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, and country-specific dummies. The definitions of each control variable can be found in the note of Table 1. is the parameter of interest, with its expected values listed in Table 2 for each financial behavior. is the participant index and is the error term. The statistical significance requirement is 5%.

Another methodological challenge is endogeneity, which is addressed using an instrumental variable (IV) methodology. This approach allows to potentially uncover a causal relationship from financial knowledge to the considered financial behaviors. We first report the logistic (or ordered logistic for categorical variables) estimations and then use IV 2SLS estimations[1] when it is required.

The IV method is used because financial knowledge may influence some financial decisions, but at the same time, financial experiences can also contribute to create financial knowledge. Fernandes et al. (2014) cited several studies and sample instruments, including the financial knowledge of the participant’s parents (Behrman et al., 2012b), which is included in the analysis. The statement for the instrument is “My parents have a good understanding of finance in general terms”, participants self-reported their agreement on a scale of one to five, where one indicates disagreement and five indicates strong agreement with the statement. This instrument is treated as a categorical variable in the estimations.

[1] OLS estimations are available in Appendix 4.

Section 5.1 presents statistics on both objective and subjective measures of financial knowledge; section 5.2 presents Logit and IV estimations to evaluate the statistical significance and sign of the relationship between each financial behavior and the objective financial knowledge measurement; section 5.3 replicates the same estimations but using the subjective measurement.

Sections 5.4 and 5.5 introduce an interaction term between the financial knowledge measures and the COVID-19 dummy variable. This interaction allows to explore the impact of the pandemic on the relationship between financial knowledge and the selected financial behaviors.

5.1 Objective and subjective financial knowledge measures

Regarding the financial knowledge measure, the percentage of people who answered all three questions correctly was 33%. This result is similar with Klapper et al. (2015), which report a percentage between 25% to 34% for the same countries. The question about compound interest rate was the most challenging, with only 42% of the participants answering it correctly. Additionally, the inflation question has the highest “do not know” responses (Table 3).

Table 3. Objective financial knowledge measure by question.

Question	Correct (%)	Incorrect (%)	Do not Know (%)
Inflation	66.65	21.56	11.79
Interest Rate	42.27	47.95	9.77
Diversification	70.29	20.54	9.17

Note: Table 3 presents the three objective financial knowledge questions and shows the percentage of participants who answered each question correctly, incorrectly, or responded “do not know”. The total number of observations was 2,834, including those who chose, “do not know”.

The self-assessment of financial knowledge is presented in Table 4, with an average score of 2.92 on a scale of one to five. The correlation between the two measurements of financial knowledge is only 0.11[1], which is in line with the findings of Walstad & Allgood (2021), who reported a correlation of 0.26. This low correlation suggests that the two measures may capture distinct components or dimensions of financial knowledge (Walstad & Allgood, 2021).

Table 4. Self-Report financial knowledge measure.

Level	Percentage (%)	Cumulative (%)
1	13.24	13.24
2	21.92	35.16
3	32.67	67.83
4	23.59	91.41
5	8.59	100

Note: The first column displays the percentage distribution for each self-assessed financial knowledge level, while the second column shows the cumulative percentage. The number of observations is 2,213.

Figure 1 depicts the percentage of individuals who answered all three objective financial knowledge questions correctly by country before and after March 2020. Peru had the highest percentage before the pandemic, while Argentina had the highest percentage after the pandemic. Except for Argentina, all the countries experienced a decline in this percentage after March 2020.

[1] The correlations tables are available in Appendix 3.

Turning to the subjective measure of financial knowledge, Figure 2 depicts the changes of this measurement for each country before and after March 2020. Similar to the objective measurement, the subjective measure decreased slightly for all the selected countries after March 2020, except for Argentina, where it increased.

5.2 Objective financial knowledge measure and financial behaviors

Table 5 presents the relationship between financial knowledge and financial behaviors using a Logit model for binary variables or an Ordered Logit model for categorical variables, estimated with clustered errors by country, as the effect of financial knowledge on financial behaviors appears to vary by each country. The results indicate that high financial knowledge is associated with having stock market investments and credit cards. Additionally, financial knowledge has a positive statistically significant correlation with common financial behaviors, such as comparing financial products before acquiring them and willingness to assume risks. On the other hand, individuals with a higher degree of financial knowledge are less likely to save money in “cadenas”.

Odds ratios in Table 5 on Logit model estimators can be obtained by applying the exponential function to the coefficients. For instance, if a participant has financial knowledge, the odds of having a credit card are almost 41.3% higher ( on average compared to someone who does not.

For those financial behaviors that have a statistically significant relationship with financial knowledge, the COVID-19 variable shows a negative association with the acquisition of financial products like a mortgage, as well as with detrimental financial behaviors such as overdraft or spending more than their income. These results could indicate that people became more cautious about their financial situation because of the pandemic.

Table 5. Binomial Logit and Ordered Logit estimations of the relations between objective financial knowledge measure and the list of financial behaviors.

	Financial Knowledge		COVID		Pseudo R-sqr
	Coefficient	SE	Coefficient	SE
Have a mortgage	-0.076	0.080	-0.274**	0.090	0.130
Have stock market investments	0.198***	0.060	-0.217	0.170	0.048
Have credit card	0.346***	0.060	-0.230	0.180	0.092
Have savings account	0.103	0.140	-0.197*	0.100	0.092
Have informal loans	-0.238	0.620	-0.133	0.190	0.050
Save in "cadenas"	-0.232***	0.070	0.006	0.070	0.021
Set aside money to save	0.092	0.080	0.263***	0.070	0.032
Financial overdraft	-0.080	0.170	-0.579***	0.080	0.089
Compare before buy	0.227**	0.080	-0.268***	0.070	0.043
Have budgets	0.098	0.070	-0.005	0.160	0.037
Financial assistance	-0.098	0.080	-0.032	0.110	0.030
Spends higher than income	-0.163	0.110	-0.390***	0.050	0.053
Consider before buy	0.019	0.150	0.102	0.070	0.004
Pay bills on time	0.242	0.130	0.157	0.260	0.040
Set long-term goals	0.016	0.100	0.070	0.080	0.008
Willingness to take risks	0.164**	0.060	-0.083	0.080	0.018

Note: The table presents the Logit and Ordered Logit estimations in the first column (coefficient) and clustered errors by country in the second column (SE) for the objective financial knowledge measure and the COVID dummy. In the last column, we report the pseudo-r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, and country-specific dummies. Statistical significance ***p<0.01, **p<0.05, *p<0.10. Ordered logit is applied to “Consider before buy”, “Pay bills on time”, “Set long-term goals”, and “Willingness to take risks” since it is a categorical variable on an integer scale of one to five.

Results in Table 5 might be endogenous, as explained in section 4.3; hence, we use the IV methodology. The endogeneity test (Durbin, 1954) (Ho: the variables are exogenous), the weak instruments test (Stock & Yogo, 2002) (Ho: weak instruments), and independent instrument tests (Sargan, 1958) (Ho: instruments are independent) are presented in Table 6. We failed to find endogeneity problems in 11 out of 16 cases, so they are omitted.

Compared to the results in Table 5, in Table 6, financial knowledge has a statistically significant negative association with having an informal loan and a negative income balance. With this method, financial knowledge has a statistically significant positive association with comparing financial products before acquiring them, considering the affordability of a purchase, and setting long-term goals. Once again, the COVID-19 parameters indicate that people were more cautious about their financial situation when the pandemic began. It is important to consider that the IV estimations are based on weak instruments in this analysis. The consequence of this issue on our results is explored at the end of section 5.

Table 6. IV estimations of the relations between objective financial knowledge measure and the list of financial behaviors.

	Financial Knowledge		COVID		Test Endogeneity	Test Weak Instrument	Test Independent Instrument
	Coefficient	SE	Coefficient	SE
Have a mortgage					No	No	Yes
Have stock market investments	0.772*	0.310	0.000	0.010	Yes	No	Yes
Have credit card					No	No	Yes
Have savings account					No	No	Yes
Have informal loans	-0.524***	0.120	-0.022	0.010	Yes	No	Yes
Save in "cadenas"					No	No	Yes
Set aside money to save					No	No	No
Financial overdraft					No	No	Yes
Compare before buy	1.479***	0.370	-0.012	0.020	Yes	No	Yes
Have budgets					No	No	No
Financial assistance					No	No	Yes
Spends higher than income	-1.245**	0.380	-0.127***	0.010	Yes	No	Yes
Consider before buy	2.936**	1.010	0.152*	0.060	Yes	No	Yes
Pay bills on time					No	No	Yes
Set long-term goals	4.939***	0.990	0.225***	0.060	Yes	No	Yes
Willingness to take risks					No	No	Yes

Note: The table presents the IV 2SLS estimations in the first column (coefficient) and clustered errors by country (SE) in the second column for the objective financial knowledge measure and the COVID dummy. In the last column, we report the r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, country-specific dummies. Statistical significance ***p<0.01, **p<0.05, *p<0.10. We test for endogeneity (Durbin, 1954) weak instruments (Stock & Yogo, 2002), and independent instruments (Sargan, 1958).

5.3 Subjective financial knowledge measure and financial behaviors

For the subjective financial knowledge measure, the estimations found more statistically significant variables than in the case of the objective measure. Out of the 16 financial behaviors, 12 exhibited statistical significance in the expected direction (Table 2), as illustrated in Table 7. The four exceptions were related to having a credit card, having a savings account, saving in “cadenas”, and incurring an overdraft, for which inferences could not find enough evidence to establish a connection with the subjective financial knowledge measure.

Similar to the results of the objective financial knowledge measure, the coefficients of the COVID-19 variable, in general, indicate that people became more cautious in their financial management after the pandemic. For instance, it increased the probability of setting money aside for savings or reduce the probability of incurring an overdraft and overspending. In contrast, behaviors such as comparing financial products or paying bills on time did not align with these increased precautions after the pandemic.

Table 7. Binomial Logit and Ordered Logit estimations of the relations between subjective financial knowledge measure and the list of financial behaviors.

	Financial Knowledge		COVID		Pseudo R-sqr
	Coefficient	SE	Coefficient	SE
Have a mortgage	0.164***	0.030	-0.262**	0.100	0.135
Have stock market investments	0.331***	0.020	-0.204	0.180	0.066
Have credit card	0.067*	0.030	-0.238	0.190	0.089
Have savings account	0.069	0.040	-0.196*	0.090	0.093
Have informal loans	-0.576***	0.060	-0.145	0.210	0.088
Save in "cadenas"	0.005	0.060	0.014	0.070	0.019
Set aside money to save	0.146***	0.010	0.272***	0.080	0.037
Financial overdraft	-0.236*	0.120	-0.604***	0.100	0.096
Compare before buy	0.425***	0.040	-0.264***	0.080	0.077
Have budgets	0.370***	0.040	0.015	0.170	0.065
Financial assistance	0.111***	0.030	-0.021	0.110	0.033
Spends higher than income	-0.245***	0.040	-0.410***	0.050	0.065
Consider before buy	0.348***	0.040	0.134*	0.060	0.020
Pay bills on time	0.283***	0.040	0.162	0.270	0.049
Set long-term goals	0.759***	0.080	0.114	0.100	0.069
Willingness to take risks	0.271***	0.040	-0.070	0.090	0.026

Note: The table presents the Logit and Ordered Logit estimations in the first column (coefficient) and clustered errors by country in the second column (SE) for the subjective financial knowledge measure and the COVID dummy. In the last column, we report the pseudo-r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, and country-specific dummies. Statistical significance ***p<0.01, **p<0.05, *p<0.10. Ordered logit is applied to “Consider before buy”, “Pay bills on time”, “Set long-term goals”, and “Willingness to take risks” since it is a categorical variable on an integer scale of one to five.

In contrast to the IV estimations for the objective measure of financial knowledge, the subjective measure has strong instruments. The results in Table 8 illustrate that having a budget has a positive relationship with self-assessed financial knowledge. On the other hand, with higher subjective financial knowledge, the probability of incurring an overdraft, or using informal loans decrease. The results also indicate that paying bills on times does not have an association with the subjective measurement of financial knowledge. These results may suggest a causal relationship between subjective financial knowledge and the presence of informal loans, incurring in overdrafts, and maintaining budgets. Still, the estimation model does not include the interaction term between financial knowledge and COVID-19 dummy.

Table 8. IV estimations of the relations between objective financial knowledge measure and the list of financial behaviors.

	Financial Knowledge		COVID		Test Endogeneity	Test Weak Instrument	Test Independent Instrument
	Coefficient	SE	Coefficient	SE
Have a mortgage					No	Yes	Yes
Have stock market investments					No	Yes	Yes
Have credit card					No	Yes	Yes
Have savings account					No	Yes	Yes
Have informal loans	-0.044***	0.010	-0.006	0.000	Yes	Yes	No
Save in "cadenas"	-0.078*	0.040	-0.003	0.010	Yes	Yes	Yes
Set aside money to save					No	Yes	No
Financial overdraft	-0.037***	0.010	-0.022***	0.000	Yes	Yes	Yes
Compare before buy					No	Yes	No
Have budgets	0.275***	0.030	0.016	0.040	Yes	Yes	Yes
Financial assistance					No	Yes	Yes
Spends higher than income					No	Yes	No
Consider before buy					No	Yes	Yes
Pay bills on time	-0.018	0.050	0.092	0.100	Yes	Yes	Yes
Set long-term goals					No	Yes	Yes
Willingness to take risks					No	Yes	Yes

Note: The table presents the IV 2SLS estimations in the first column (coefficient) and clustered errors by country (SE) in the second column for the subjective financial knowledge measure and the COVID dummy. In the last column, we report the r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, and dummies by country. Statistical significance ***p<0.01, **p<0.05, *p<0.10. We test for endogeneity (Durbin, 1954) weak instruments (Stock & Yogo, 2002), and independent instruments (Sargan, 1958).

In summary, the variety in the results and the low correlation between both financial knowledge measurements indicates that each one helps capture a different dimension of an individual’s financial knowledge (Walstad & Allgood, 2021). The behaviors for which we found sufficient evidence to be positively related with the subjective measure include having a mortgage, setting money aside to save, budgeting, seeking for financial assistance, considering the affordability of a purchase, and setting long-term goals. Conversely, having informal loans, incurring overdrafts, and overspending are negatively associated with the subjective financial knowledge measure.

The objective financial knowledge measure is positively associated with having a credit card and negatively with saving in a “cadena”. We also found statistically significant positive relationships with both financial knowledge measures with having stock market investments, comparing financial products before acquiring them, and willingness to assume risk.

We could not find a significant correlation with having a saving account and paying bills on time for both of the financial knowledge measures. Even though we observed the expected sign for having a savings account and paying bills, the lack of statistically significant might be associated with the selection bias in our sample, where we expect the participants to have more financial knowledge than the general population, making those decisions appear less substantial.

5.4 COVID-19 impact on the relationship between the objective financial knowledge measure and financial behaviors

The effect of the financial knowledge on financial behaviors might be affected by the pandemic. For instance, it is possible that even if your financial knowledge is high, your capacity to pay bills on time or be accepted for a mortgage might be null due to a negative income shock caused by the pandemic. This section presents the same estimations made in section 5.2, but an interaction term between the objective financial knowledge measure and the COVID-19 indicator is included.

Compared with Table 5, in Table 9, the results indicate that the COVID-19 indicator modulates the relationship between the objective financial knowledge measure and some financial behaviors. Specifically, seeking financial assistance shows a positive statistical significance relation with the interaction term, suggesting that the pandemic period is related to an increased in the probability of seeking financial assistance for financially knowledgeable participants. A similar pattern is observed for informal loans, which has an unexpected sign. This unexpected sign for the informal loans in the interaction term is corrected in the IV estimations.

In contrast, financial overdrafts have a negative correlation with the interaction term, suggesting that the probability of overdraft was lower during the pandemic period for financially knowledgeable participants. The results for the other behaviors do not indicate a relevant role for the interaction term.

Table 9. Binomial Logit and Ordered Logit estimations of the relations between objective financial knowledge measure, the list of financial behaviors and the COVID-19 pandemic.

	Financial Knowledge		FK*COVID		COVID		Pseudo R-sqr
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Have a mortgage	-0.365	0.200	0.393*	0.190	-0.417***	0.090	0.131
Have stock market investments	0.074	0.170	0.166	0.250	-0.284*	0.140	0.048
Have credit card	0.234	0.150	0.144	0.170	-0.278	0.160	0.092
Have savings account	0.149	0.560	-0.059	0.540	-0.178	0.270	0.092
Have informal loans	-0.573	0.660	0.442**	0.150	-0.243	0.220	0.051
Save in "cadenas"	-0.507*	0.230	0.356	0.320	-0.103	0.070	0.021
Set aside money to save	0.253***	0.070	-0.209	0.140	0.339***	0.060	0.033
Financial overdraft	0.193	0.150	-0.406**	0.140	-0.417***	0.110	0.090
Compare before buy	-0.083	0.200	0.402	0.310	-0.406***	0.100	0.044
Have budgets	0.004	0.170	0.123	0.170	-0.048	0.120	0.037
Financial assistance	-0.534***	0.160	0.568***	0.150	-0.239	0.150	0.033
Spends higher than income	-0.177	0.100	0.018	0.120	-0.397***	0.030	0.053
Consider before buy	0.264***	0.070	-0.324	0.270	0.215	0.150	0.005
Pay bills on time	0.248*	0.100	-0.007	0.130	0.159	0.220	0.040
Set long-term goals	-0.056	0.110	0.094	0.090	0.036	0.050	0.008
Willingness to take risks	0.350*	0.150	-0.242	0.200	0.001	0.050	0.018

Note: The table presents the Logit and Ordered Logit estimations in the first column (coefficient) and clustered errors by country (SE) in the second column for the objective financial knowledge measure, its interaction with the COVID dummy, and the COVID dummy. In the last column, we report the pseudo-r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, and country-specific dummies. Statistical significance ***p<0.01, **p<0.05, *p<0.10. Ordered logit is applied to “Consider before buy”, “Pay bills on time”, “Set long-term goals”, and “Willingness to take risks” since it is a categorical variable on an integer scale of one to five.

The results using the IV method[1] (see table 10) are similar to the IV estimators without the interaction term in Table 6, with the difference being that the statistical significance shift form the financial knowledge variable to the interaction term. Hence, the COVID-19 period is associated with a reduction in the probability of having an informal loan and experiencing a deficit income for financially knowledgeable participants. In contrast, the COVID-19 period is associated with an increase in the the probability of comparing financial products before acquiring them, having a budget, considering the affordability of a purchase, and setting long-term goals for financially knowledgeable participants. This evidence suggests that the expected effect of the financial knowledge on financial behaviors was modulated by the COVID-19 period, indicating that a financially knowledgeable person was better equipped to cope with the financial challenges of the pandemic period.

Again, these estimations are based on weak instruments, making it challenging to establish and adequate causal conclusion. Nevertheless, they provide reference for comparing with the other results and for future studies.

Table 10. IV estimations of the relations between objective financial knowledge measure, the list of financial behaviors and the COVID-19 pandemic.

	Financial Knowledge		FK*COVID		COVID		Test Endogeneity	Test Weak Instrument	Test Independent Instrument
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Have a mortgage							No	No	Yes
Have stock market investments							No	No	No
Have credit card							No	No	Yes
Have savings account							No	No	Yes
Have informal loans	-0.052	0.060	-0.466***	0.120	0.146***	0.030	Yes	No	Yes
Save in "cadenas"							No	No	Yes
Set aside money to save							No	No	Yes
Financial overdraft							No	No	Yes
Compare before buy	0.178	0.420	1.739**	0.620	-0.623**	0.220	Yes	No	Yes
Have budgets	-0.330*	0.140	3.202***	0.360	-1.054***	0.130	Yes	No	Yes
Financial assistance							No	No	Yes
Spends higher than income	0.055	0.180	-1.548***	0.170	0.424***	0.070	Yes	No	Yes
Consider before buy	0.083	0.390	3.487**	1.340	-1.082*	0.500	Yes	No	Yes
Pay bills on time							No	No	Yes
Set long-term goals	-0.624	0.710	7.260***	1.690	-2.331***	0.490	Yes	No	Yes
Willingness to take risks							No	No	Yes

Note: The table presents the IV 2SLS estimations in the first column (coefficient) and clustered errors by country (SE) in the second column for the objective financial knowledge measure, its interaction with the COVID dummy, and the COVID dummy. In the last column, we report the r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, and dummies by country. Statistical significance ***p<0.01, **p<0.05, *p<0.10. We test for endogeneity (Durbin, 1954) weak instruments (Stock & Yogo, 2002), and independent instruments (Sargan, 1958).

5.5 COVID-19 impact on the relationship between the subjective financial knowledge measure and financial behaviors

The previous exercise is also conducted for the subjective financial knowledge measure. In Table 11, where the interaction term is included, the results are similar to those in Table 7, where 14 behaviors exhibit statistical significance. In contrast to the results of the objective financial knowledge measure (Table 9), the interaction term is only statistically significant for having a credit card. This estimation is positive; hence, an increase of the probability of having a credit card is associated with the pandemic period for participants that self-assessed their financial knowledge as high.

Table 11. Binomial Logit and Ordered Logit estimations of the relations between subjective financial knowledge measure, the list of financial behaviors and the COVID-19 pandemic.

	Financial Knowledge		FK*COVID		COVID		Pseudo R-sqr
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Have a mortgage	0.164***	0.030	0.012	0.060	-0.266*	0.110	0.135
Have stock market investments	0.327***	0.020	0.194	0.110	-0.277	0.150	0.067
Have credit card	0.058*	0.030	0.364***	0.080	-0.351	0.190	0.092
Have savings account	0.067	0.040	0.076*	0.030	-0.220*	0.090	0.093
Have informal loans	-0.575***	0.050	-0.009	0.580	-0.142	0.340	0.088
Save in "cadenas"	0.009	0.060	-0.144	0.110	0.060	0.070	0.020
Set aside money to save	0.146***	0.010	0.014	0.110	0.268***	0.060	0.037
Financial overdraft	-0.233*	0.120	-0.187	0.170	-0.540***	0.140	0.096
Compare before buy	0.420***	0.040	0.259	0.140	-0.346***	0.090	0.079
Have budgets	0.369***	0.040	0.065	0.070	-0.006	0.150	0.065
Financial assistance	0.110***	0.030	0.024	0.070	-0.029	0.110	0.033
Spends higher than income	-0.243***	0.040	-0.115	0.120	-0.371***	0.030	0.065
Consider before buy	0.350***	0.040	-0.112	0.210	0.171	0.120	0.021
Pay bills on time	0.279***	0.040	0.194	0.150	0.103	0.220	0.050
Set long-term goals	0.762***	0.080	-0.101	0.120	0.148	0.080	0.069
Willingness to take risks	0.269***	0.040	0.059	0.070	-0.089	0.080	0.026

Note: The table presents the Logit and Ordered Logit estimations in the first column (coefficient) and clustered errors by country (SE) in the second column for the subjective financial knowledge measure, its interaction with the COVID dummy, and the COVID dummy. In the last column, we report the pseudo-r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, urban area, and country-specific dummies. Statistical significance ***p<0.01, **p<0.05, *p<0.10. Ordered logit is applied to “Consider before buy”, “Pay bills on time”, “Set long-term goals”, and “Willingness to take risks” since it is a categorical variable on an integer scale of one to five.

IV estimations with the interaction term are shown in Table 12. Again, the interaction term does not have a statistical significance with any financial behaviors. However, the subjective financial knowledge measure positively relates to whether the participants have budgets and negatively relates to having informal loans. The results also indicate that paying bills on times does not have an association with the subjective measurement of financial knowledge, a result consistent from the Table 8. These estimations are based on weak instruments.

Table 12. IV estimations of the relations between subjective financial knowledge measure, a list of financial behaviors and the COVID-19 pandemic.

	Financial Knowledge		FK*COVID		COVID		Test Endogeneity	Test Weak Instrument	Test Independent Instrument
	Coefficient	SE	Coefficient	SE	Coefficient	SE
Have a mortgage							No	No	Yes
Have stock market investments							No	No	No
Have credit card							No	No	Yes
Have savings account							No	No	Yes
Have informal loans	-0.042***	0.010	-0.106	0.100	0.028	0.040	Yes	No	Yes
Save in "cadenas"	-0.073	0.050	-0.188	0.270	0.059	0.090	Yes	No	Yes
Set aside money to save							No	No	Yes
Financial overdraft							No	No	Yes
Compare before buy							No	No	Yes
Have budgets	0.226***	0.040	0.806	0.420	-0.249	0.140	Yes	No	Yes
Financial assistance							No	No	Yes
Spends higher than income							No	No	Yes
Consider before buy							No	No	Yes
Pay bills on time	-0.089	0.070	1.890	0.970	-0.527	0.330	Yes	No	Yes
Set long-term goals							No	No	Yes
Willingness to take risks							No	No	Yes

Note: The table presents the IV 2SLS estimations in the first column (coefficient) and clustered errors by country (SE) in the second column for the subjective financial knowledge measure, its interaction with the COVID dummy, and the COVID dummy. In the last column, we report the r-squared. The estimations include the control variables of age, education level, sex, marital status, income, number of adults, number of children, stable income, rural area, and country-specific dummies. Statistical significance ***p<0.01, **p<0.05, *p<0.10. We test for endogeneity (Durbin, 1954) weak instruments (Stock & Yogo, 2002), and independent instruments (Sargan, 1958).

To summarize, the IV results presented in Table 6, Table 10, and Table 12 are based on weak instruments, creating an estimation bias towards OLS estimations and affect the validity of statistical inference (Andrews et al., 2019; Hahn & Hausman, 2003). However, our sample is large, and we attempted to use alternative instruments, such as the sibling’s subjective financial knowledge level and the participants’ academic level. Unfortunately, these alternatives did not lead to improvements in the tests results and the estimation. As a result, the results from these tables are only indicative and complementary and should not be used as evidence of a causal relationship.

Despite this issue of weak instruments and the selection bias, our results highlight two crucial findings for futures research and financial knowledge programs, especially in Latin America: i) participants with higher financial knowledge coped better during the pandemic and exhibited better financial practices, ii) our results provide a reference estimations to understand the relationships between 16 financial behaviors and two financial knowledge measures in Latin American countries, and iii) the causal relationship between financial knowledge and certain financial behaviors is a significant issue that has received limited attention in studies focusing on for Latin America countries. There is an empirical need to find valid instruments and address the endogeneity problem for non-experimental and non-qualitative studies.

[1] The endogenous variables are the financial knowledge measures and the interaction term. The instruments are the parents’ financial knowledge and the interaction of the parents’ financial knowledge with the COVID-19 indicator.

Results align with the findings of previous studies listed in Table 2 regarding the correlation between financial knowledge and financial behaviors. Although, there are two exceptions in the cases of having a savings account and paying bills on time, where the evidence was not enough to establish a relationship. This lack of evidence could be attributed to sample bias, as participants in our study may possess higher financial knowledge than the population, making these financial decisions seem trivial and less relevant.

Both the objective and subjective knowledge measures exhibit positive correlations with behaviors such as for investing in the stock market, having a credit card, comparing financial product before acquiring them, seeking financial assistance, and willingness to take risks. The objective measure shows a negative relation with saving in “cadenas” and incurring in overdrafts, while the subjective measure has a positive relationship with setting aside money to save, budgeting, considering affordability before making purchases, and setting long term goals.

Some of those relationships are influenced by the COVID-19 pandemic, affecting the probability of having a credit card, incur in a financial overdraft, and seeking financial assistance. In general, the COVID-19 period led to more precautionary financial behaviors, particularly among participants with higher financial knowledge.

In the case of having informal loans, our results contradict expectations, as the sign of the interaction term is positive using the objective measure. This suggest that during the pandemic, the probability of having an informal loan was higher for people with high financial knowledge. This evidence supports the arguments made by Willis (2011) and Huston (2010), which explain that financial knowledge does not necessarily lead to improvements in the financial well-being, as the last requires individuals to make the correct decisions that are not guarantee by the knowledge dimension.

As a consequence, for future research in Latin America, we recommend to clearly identifying the difference between financial literacy and financial knowledge. Huston's (2010) frameworks is insightful, as it indicates that “financial literacy consists of both knowledge and application of human capital specific to personal finance”. This study utilize questions similar to the “The Big Three”, which primarily assess what people know about finance, rather than how they manage their personal finance issues. This distinction is often overlooked in financial literacy studies in Latin America.

Regarding our methodology, some results may be affected by weak instruments. The issue of endogeneity complicates the assessment of causal relationships in empirical research. Therefore, it is important to develop educational programs that can evaluate the impact of financial behaviors to confirm our findings. Additionally, the instrument variables used in other regions (Fernandes et al., 2014) do not perform adequately in this Latin America sample, supporting the idea of Santini et al. (2019) that the impact of financial literacy varies between countries, even more significantly, between continents.

Following the perspective of Poon & Olen (2015), the lack of financial literacy cannot be resolved solely through understanding concepts, there is an opportunity for future studies to explore the complexity of financial literacy and its dependency on factors like culture and gender. With this deeper understanding, more effective educational programs can be developed to improve the financial well-being of participants, especially in Latin America, where studies on financial literacy remain relatively low (Goyal & Kumar, 2021).

Worldwide, governments have been implementing financial education programs mainly to improve financial access and consumer protection (Atkinson, A.; Messy, 2013); however, the impact of those programs on people’s financial behaviors remains inconclusive (Fernandes et al., 2014b; Miller et al., 2015). Furthermore, despite all the numerous publications on financial knowledge, most have ignored the endogeneity problem between financial knowledge and financial behaviors, especially in Latin America (Prado et al., 2022).

This study collected data from Argentina, Colombia, Ecuador, Mexico, and Peru and measured financial knowledge objectively and subjectively to address measurement errors. The study estimated the relationship between financial knowledge and 16 financial behaviors using logit (or ordered logistic) and IV methodologies to address the endogeneity issue. The sample has selection bias; hence the results have limitations in generalization. Nevertheless, the findings can serve as a reference to understand the role of financial knowledge in people’s financial behaviors in Latin America. Additionally, the proposed set of instruments is weak, despite using instruments from previous studies (Fernandes et al., 2014), making it difficult to assess causal relationships.

The results indicate that between the objective and subjective financial knowledge measures have lower correlations, suggesting that each measures a different dimension of financial knowledge (Walstad & Allgood, 2021). Moreover, 14 out of the 16 considered behaviors show a statistically significant relationship with at least one of the two financial knowledge measures. All of theme have the expected sign (Table 2), except for informal loans, which has a positive relationship with the objective financial knowledge measure. This result might support the idea that financial knowledge does not necessarily lead to improved our financial decisions (Huston, 2010; Willis, 2011). Results did not find statistically significant relationships between both financial knowledge measures and having a savings account or paying bills on time.

In general, during the pandemic period, participants were more careful with the management of their financial products. This effect is higher among participants with a high of objective financial knowledge. Our results also suggest that the subjective measure does not interact with the pandemic factor. These findings indicate that participants with greater financial knowledge exhibit more appropriate financial behaviors during the pandemic compared to participants with less financial knowledge.

Future research and projects aiming to improve the financial well-being of families, in particular in Latin America, should consider that financial knowledge may not explain all financial behaviors. Consequently, future financial education programs should include impact evaluations and specific objectives to assist families with their financial issues. Although financial education is crucial for improving financial decisions, it is hardly the sole factor influencing families to make better financial decisions.

Conflict of Interest: The authors have no conflict of interest that might be perceived to influence the results and/or discussion reported in this paper.

Data availability: The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Atkinson, A.; Messy, F. (2013). Promoting financial inclusion through financial education: Andrews, I., Stock, J. H., & Sun, L. (2019). Weak Instruments in Instrumental Variables Regression: Theory and Practice. Annual Review of Economics, 11, 727–753. https://doi.org/10.1146/annurev-economics-080218-025643

Dew, J., & Xiao, J. J. (2011). The financial management behavior scale: Development and validation. Journal of Financial Counseling and Planning, 22(1), 43–59.

Fernandes, D., Lynch, J. G., & Netemeyer, R. G. (2014). Financial literacy, financial education, and downstream financial behaviors. Management Science, 60(8), 1861–1883. https://doi.org/10.1287/mnsc.2013.1849

Goyal, K., & Kumar, S. (2021). Financial literacy: A systematic review and bibliometric analysis. International Journal of Consumer Studies, 45(1), 80–105. https://doi.org/10.1111/ijcs.12605

Hahn, J., & Hausman, J. (2003). American Economic Association Weak Instruments : Diagnosis and Cures in Empirical Econometrics Author ( s ): Jinyong Hahn and Jerry Hausman Source : The American Economic Review , Vol . 93 , No . 2 , Papers and Proceedings of the One Hundred Fifteenth Ann. American Economic Association, 93(2), 118–125.

Huston, S. J. (2010). Measuring Financial Literacy. Journal of Consumer Affairs, 44(2), 296–316. https://doi.org/10.1111/j.1745-6606.2010.01170.x

Kaiser, T., & Menkhoff, L. (2017). Does financial education impact financial literacy and financial behavior, and if so, when? World Bank Economic Review, 31(3), 611–630. https://doi.org/10.1093/wber/lhx018

Kaiser, T., & Menkhoff, L. (2020). Financial education in schools: A meta-analysis of experimental studies. Economics of Education Review, 78(November 2018), 101930. https://doi.org/10.1016/j.econedurev.2019.101930

Knoll, M. A. Z., & Houts, C. R. (2012). The Financial Knowledge Scale: An Application of Item Response Theory to the Assessment of Financial Literacy. Journal of Consumer Affairs, 46(3), 381–410. https://doi.org/10.1111/j.1745-6606.2012.01241.x

Poon, M., & Olen, H. (2015). Does literacy improve finance? Public Understanding of Science, 24(3), 272–284. https://doi.org/10.1177/0963662515571209

Ranyard, R., McNair, S., Nicolini, G., & Duxbury, D. (2020). An item response theory approach to constructing and evaluating brief and in-depth financial literacy scales. Journal of Consumer Affairs, 54(3), 1121–1156. https://doi.org/10.1111/joca.12322

Remund, D. L. (2010). Financial literacy explicated: The case for a clearer definition in an increasingly complex economy. Journal of Consumer Affairs, 44(2), 276–295. https://doi.org/10.1111/j.1745-6606.2010.01169.x

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(2), 243–262. https://doi.org/10.1111/joca.12010

Walstad, W. B., & Allgood, S. (2021). The likely influence of financial literacy on financial behaviors. In The Routledge Handbook of Financial Literacy (pp. 153–170). Routledge.

Willis, L. E. (2011). The financial education fallacy. American Economic Review, 101(3), 429–434. https://doi.org/10.1257/aer.101.3.429

OECD/INFE evidence, policies, and practice. OECD Publishing, 34. https://doi.org/10.1787/5k3xz6m88smp-en

Badarinza, C., Campbell, J. Y., & Ramadorai, T. (2016). International Comparative Household Finance. Annual Review of Economics, 8, 111–144. https://doi.org/10.1146/annurev-economics-080315-015425

Behrman, J. R., Mitchell, O. S., Soo, C., & Bravo, D. (2012a). Financial Literacy, Schooling, and Wealth Accumulation. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.1688985

Behrman, J. R., Mitchell, O. S., Soo, C. K., & Bravo, D. (2012b). How financial literacy affects household wealth accumulation. American Economic Review, 102(3), 300–304. https://doi.org/10.1257/aer.102.3.300

Bernal, N., & Olivera, J. (2020). Choice of pension management fees and effects on pension wealth. Journal of Economic Behavior and Organization, 176, 539–568. https://doi.org/10.1016/j.jebo.2020.03.036

Brüggen, E. C., Hogreve, J., Holmlund, M., Kabadayi, S., & Löfgren, M. (2017). Financial well-being: A conceptualization and research agenda. Journal of Business Research, 79, 228–237. https://doi.org/10.1016/j.jbusres.2017.03.013

Campbell, J. Y. (2006). Household finance. Journal of Finance, 61(4), 1553–1604. https://doi.org/10.1111/j.1540-6261.2006.00883.x

Cao-Alvira, J. J., Novoa-Hoyos, A., & Núñez-Torres, A. (2021). On the financial literacy, indebtedness, and wealth of Colombian households. Review of Development Economics, 25(2), 978–993. https://doi.org/10.1111/rode.12739

Collins, J. M. (2012). Financial Advice: A Substitute for Financial Literacy? SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2046227

de la Fuente-Mella, H., Umaña-Hermosilla, B., Fonseca-Fuentes, M., & Elórtegui-Gómez, C. (2021). Multinomial logistic regression to estimate the financial education and financial knowledge of university students in Chile. Information (Switzerland), 12(9). https://doi.org/10.3390/info12090379

Durbin, J. (1954). Interview Bias Author ( s ): C . A . Moser Source : Revue de l ’ Institut International de Statistique / Review of the International Published by : International Statistical Institute ( ISI ) Stable URL : https://www.jstor.org/stable/1401500. Review of the International Statistical Institute, 22(1), 23–32.

Fernandes, D., Lynch, J. G., & Netemeyer, R. G. (2014a). Financial literacy, financial education, and downstream financial behaviors. Management Science, 60(8), 1861–1883. https://doi.org/10.1287/mnsc.2013.1849

Fernandes, D., Lynch, J. G., & Netemeyer, R. G. (2014b). Financial literacy, financial education, and downstream financial behaviors. Management Science, 60(8), 1861–1883. https://doi.org/10.1287/mnsc.2013.1849

Frisancho, V. (2016). How to Raise Household Savings in LAC: Constraints and Best Practices. April.

García Mata, O. (2021). The effect of financial literacy and gender on retirement planning among young adults. International Journal of Bank Marketing, 39(7), 1068–1090. https://doi.org/10.1108/IJBM-10-2020-0518

Gerardi, K., Goette, L., & Meier, S. (2013). Numerical ability predicts mortgage default. Proceedings of the National Academy of Sciences of the United States of America, 110(28), 11267–11271. https://doi.org/10.1073/pnas.1220568110

Gomes, F., Haliassos, M., & Ramadorai, T. (2021). Household finance. Journal of Economic Literature, 59(3), 919–1000. https://doi.org/10.1257/JEL.20201461

Goyal, K., & Kumar, S. (2021). Financial literacy: A systematic review and bibliometric analysis. International Journal of Consumer Studies, 45(1), 80–105. https://doi.org/10.1111/ijcs.12605

Guiso, L., & Sodini, P. (2013). Household Finance: An Emerging Field. In Handbook of the Economics of Finance (Vol. 2, Issue PB). Elsevier B.V. https://doi.org/10.1016/B978-0-44-459406-8.00021-4

Gutter, M., & Copur, Z. (2011). Financial Behaviors and Financial Well-Being of College Students: Evidence from a National Survey. Journal of Family and Economic Issues, 32(4), 699–714. https://doi.org/10.1007/s10834-011-9255-2

Hastings, J. S., Madrian, B. C., & Skimmyhorn, W. L. (2013). Financial literacy, financial education, and economic outcomes. Annual Review of Economics, 5, 347–373. https://doi.org/10.1146/annurev-economics-082312-125807

Hernández-Mejía, S., García-Santillán, A., & Moreno-García, E. (2021). Financial literacy and the use of credit cards in Mexico. Journal of International Studies, 14(4), 97–112. https://doi.org/10.14254/2071-8330.2021/14-4/7

Jappelli, T., & Padula, M. (2013). Investment in financial literacy and saving decisions. Journal of Banking and Finance, 37(8), 2779–2792. https://doi.org/10.1016/j.jbankfin.2013.03.019

Karakurum-Ozdemir, K., Kokkizil, M., & Uysal, G. (2019). Financial Literacy in Developing Countries. Social Indicators Research, 143(1), 325–353. https://doi.org/10.1007/s11205-018-1952-x

Kim, K. T., Anderson, S. G., & Seay, M. C. (2019). Financial Knowledge and Short-Term and Long-Term Financial Behaviors of Millennials in the United States. Journal of Family and Economic Issues, 40(2), 194–208. https://doi.org/10.1007/s10834-018-9595-2

Klapper, L., Lusardi, A., & Panos, G. A. (2013). Financial literacy and its consequences: Evidence from Russia during the financial crisis. Journal of Banking and Finance, 37(10), 3904–3923. https://doi.org/10.1016/j.jbankfin.2013.07.014

Klapper, L., Lusardi, A., & van Oudheusden, P. (2015). Financial literacy around the world. World Bank. Washington DC: World Bank.

Lin, C., Hsiao, Y. J., & Yeh, C. Y. (2017). Financial literacy, financial advisors, and information sources on demand for life insurance. Pacific Basin Finance Journal, 43(April), 218–237. https://doi.org/10.1016/j.pacfin.2017.04.002

Literacy, U. S. F., & Commission, E. (2020). US national strategy for financial literacy 2020. Washington, DC: US Financial Literacy and Education Commission. September.

Lusardi, A., & Mitchell, O. S. (2007). Financial Literacy and Retirement Prepared-ness: Evidence and Implications for Financial Education. Business Economics, 42, January, 35–44. https://link.springer.com/content/pdf/10.2145%2F20070104.pdf

Lusardi, A., & Mitchell, O. S. (2014). The economic importance of financial literacy: Theory and evidence. Journal of Economic Literature, 52(1), 5–44. https://doi.org/10.1257/jel.52.1.5

Lusardi, A., & Mitchell, O. S. (2017). How Ordinary Consumers Make Complex Economic Decisions: Financial Literacy and Retirement Readiness. Quarterly Journal of Finance, 7(3), 1–31. https://doi.org/10.1142/S2010139217500082

Lusardi, A., & Tufano, P. (2015). Debt literacy, financial experiences, and overindebtedness. In Journal of Pension Economics and Finance (Vol. 14, Issue 4). https://doi.org/10.1017/S1474747215000232

Martínez A, C., & Puentes, E. (2018). Micro-entrepreneurship Debt Level and Access to Credit: Short-Term Impacts of a Financial Literacy Program. European Journal of Development Research, 30(4), 613–629. https://doi.org/10.1057/s41287-017-0085-4

Méndez Prado, S. M., Zambrano Franco, M. J., Zambrano Zapata, S. G., Chiluiza García, K. M., Everaert, P., & Valcke, M. (2022). A Systematic Review of Financial Literacy Research in Latin America and The Caribbean. Sustainability, 14(7), 3814. https://doi.org/10.3390/su14073814

Miller, M., Reichelstein, J., Salas, C., & Zia, B. (2015). Can you help someone become financially capable? A meta-analysis of the literature. World Bank Research Observer, 30(2), 220–246. https://doi.org/10.1093/wbro/lkv009

Mitchell, O. S., & Lusardi, A. (2015). Financial Literacy and Economic Outcomes: Evidence and Policy Implications . The Journal of Retirement, 3(1), 107–114. https://doi.org/10.3905/jor.2015.3.1.107

Muñoz-Murillo, M., Álvarez-Franco, P. B., & Restrepo-Tobón, D. A. (2020). The role of cognitive abilities on financial literacy: New experimental evidence. Journal of Behavioral and Experimental Economics , 84(October 2019), 101482. https://doi.org/10.1016/j.socec.2019.101482

Nicolini, G., & Cude, B. J. (2022). The Routledge Handbook of Financial Literacy. Routledge.

Prado, S. M. M., Everaert, P., & Valcke, M. (2019). The impact of video-vignettes to enhance the financial literacy level of Ecuadorian university students. ACM International Conference Proceeding Series, 300–303. https://doi.org/10.1145/3369255.3369273

Santini, F. D. O., Ladeira, W. J., Mette, F. M. B., & Ponchio, M. C. (2019). The antecedents and consequences of financial literacy: a meta-analysis. International Journal of Bank Marketing, 37(6), 1462–1479. https://doi.org/10.1108/IJBM-10-2018-0281

Santos, D. B., Mendes-Da-Silva, W., & Gonzalez, L. (2018). Lower financial literacy induces use of informal loans. RAE Revista de Administracao de Empresas, 58(1), 44–59.

Sargan, J. D. (1958). The estimation of economic relationships using instrumental variables. Econometrica: Journal of the Econometric Society, 393–415.

Stock, J. H., & Yogo, M. (2002). Testing for weak instruments in linear IV regression. National Bureau of Economic Research Cambridge, Mass., USA.

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(4), 593–608. https://doi.org/10.1016/j.joep.2011.02.004

van Rooij, M. C. J., Lusardi, A., & Alessie, R. J. M. (2012). Financial Literacy, Retirement Planning and Household Wealth. Economic Journal, 122(560), 449–478. https://doi.org/10.1111/j.1468-0297.2012.02501.x

van Rooij, M., Lusardi, A., & Alessie, R. (2011). Financial literacy and stock market participation. Journal of Financial Economics, 101(2), 449–472. https://doi.org/10.1016/j.jfineco.2011.03.006

Vieira, K. M., Potrich, A. C. G., & Mendes-Da-Silva, W. (2019). A financial literacy model for university students. In Individual Behaviors and Technologies for Financial Innovations (pp. 69–95). Springer.

von Gaudecker, H. M. (2015). How Does Household Portfolio Diversification Vary with Financial Literacy and Financial Advice? Journal of Finance, 70(2), 489–507. https://doi.org/10.1111/jofi.12231

Walstad, W. B., & Allgood, S. (2021). The likely influence of financial literacy on financial behaviors. In The Routledge Handbook of Financial Literacy (pp. 153–170). Routledge.

No competing interests reported.

Download PDF

Version 2

posted

You are reading this older preprint version

Read the latest preprint version →

Financial Knowledge and Finance Behaviors in Latin America and its Relationship with the COVID-19 Pandemic Period

Status:

Version 2

Abstract

Figures

Introduction

Literature Review

Data

Methodology

Results

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 2