1. Introduction
Introduced by Nakamoto (Reference Nakamoto2008), Bitcoin was designed to function as a peer-to-peer electronic cash system outside the system established by central banks and intermediary financial institutions. Bitcoin was touted as both a payment method and a financial asset, attracting considerable attention and investment worldwide. Footnote 1 However, due to its inherent volatility and inefficiency as a payment method, Bitcoin has evolved into more of an investment product rather than a reliable means of payment (Balutel et al. Reference Balutel, Engert, Henry, Huynh and Voia2024; Henry et al. Reference Henry, Huynh and Nicholls2019a; Stix Reference Stix2021).
The Bank of Canada monitors and conducts research on Bitcoin and other cryptoassets for several reasons. One primary motivation is the exploration of these digital currencies in the context of potentially issuing a central bank digital currency (CBDC). This exploration necessitates careful consideration of two crucial conditions outlined in Lane (Reference Lane2020, Reference Lane2021) that would lead the Bank to consider issuing a CBDC: if cash could no longer be used for a wide range of transactions or if private digital currencies make serious inroads as an alternative payment method. Further, the Bank of Canada’s Financial System Review (2022) finds that: “Cryptoasset markets continue to evolve and grow rapidly, and price volatility remains high. While they do not yet pose a systemic risk to the Canadian financial system, the lack of a regulatory framework means they operate without many of the safeguards that exist in the traditional financial system. This exposes investors to risks such as large and sudden financial losses due to fraud, price declines, or a run on stablecoins.”
Balutel et al. (Reference Balutel, Engert, Henry, Huynh and Voia2024) find evidence from survey data that in 2021, 25 percent of cryptoasset owners experienced price crashes. Other incidents reported include lost access to a wallet (11%), initial coin offering scams (7%), stolen funds (7%), and data breaches (6%). Footnote 2
The Bank of Canada’s Financial System Review (2023) adds that: “Cryptoasset markets do not currently represent a significant concern for the stability of the Canadian financial system. They remain small and mostly separate from the financial system. If they do become more interconnected, shocks in these markets could spread to the broader financial system and affect financial stability.”
In July 2023, Canada’s banking and insurance regulator (OSFI) set out its regulatory requirements for banks’ and insurers’ exposures to cryptoassets, to come into effect in 2025, replacing the interim OSFI advisory on cryptoassets that was released in 2022 (Office of the Superintendent of Financial Institutions 2023a, 2023b).
While constituting a relatively small market, understanding what drives the adoption of cryptoassets and the demographic profiles of their owners is valuable. The percentage of Canadians who own Bitcoin increased significantly from 5 percent in 2018–2020 to 13 percent in 2021 (Balutel et al., Reference Balutel, Engert, Henry, Huynh and Voia2022a, Reference Balutel, Engert, Henry, Huynh and Voia2024), with the highest concentration observed among men. Footnote 3 The surge in media attention and ease of access to cryptoassets, driven by the prevailing fear of missing out (FoMO), has likely influenced individuals who typically avoid high-risk investments to participate. As highlighted by Gerrans et al. (Reference Gerrans, Abisekaraj and Liu2023), FoMO is a widely acknowledged motivator for cryptocurrency investment. Therefore, it is important to have a solid foundation in crypto and financial literacy, as it can help individuals avoid impulsive decisions and cope with the risks and uncertainties in the dynamic world of cryptoassets.
Crypto literacy, as a proxy for digital literacy, is essential for investors to help them understand and navigate their engagement with such assets. In a recent survey, Ontario Securities Commission (2022) found that while 51 percent of Canadians knew the correct definition of cryptoassets, their average score on a related knowledge test was only 37 percent, indicating limited understanding of practical, legal, and regulatory aspects. Another U.S. study highlights a significant gap in crypto literacy, revealing that 91 percent of participants failed a crypto literacy test despite increased awareness fueled by media attention. Footnote 4 Additionally, Bannier et al. (Reference Bannier, Meyll, Röder and Walter2019) showed that respondents could correctly answer only 3 out of 6 crypto-related questions on average. Footnote 5 Their research also found a gender gap in crypto literacy and estimated that financial literacy alone accounted for approximately 40 percent of this observed gender gap in understanding characteristics of Bitcoin.
Financial literacy is another key characteristic of interest for this paper. Studies suggest that cryptoasset owners in Japan (Fujiki Reference Fujiki2020) and Austria (Stix Reference Stix2021) are more financially literate than non-owners. In contrast, Bitcoin ownership in Canada has historically been more common among those with low financial literacy, although there was an increased share of those with high financial literacy in 2021 (Balutel et al., Reference Balutel, Engert, Henry, Huynh and Voia2022a, Reference Balutel, Engert, Henry, Huynh and Voia2024). Further, Fujiki (Reference Fujiki2021) explored the heterogeneity of Japanese cryptoasset owners and found that those owners without investment experience in risky conventional assets display lower levels of financial literacy relative to both cryptoasset owners and non-owners with investment experience. In the context of the gender gap in financial literacy, a substantial body of literature emphasizes the widely recognized disparity in financial literacy between genders. This discrepancy carries significant implications for participation in financial markets and overall financial well-being (Bucher-Koenen et al. Reference Bucher-Koenen, Lusardi and Alessie2017, Reference Bucher-Koenen, Alessie, Lusardi and Van Rooij2021; Lusardi and Mitchell Reference Lusardi and Mitchell2011b, Reference Lusardi and Mitchell2011c; van Rooij et al. Reference van Rooij, Lusardi and Alessie2011). Footnote 6
Some proponents argue that blockchain technology has the potential to mitigate the digital divide and enhance financial inclusion, as indicated by Hydary (Reference Hydary2019). However, Carmona (Reference Carmona2022) contends that the anticipated benefits of cryptoassets for financial inclusion have not materialized. This represents a divergence of views on the actual impact of blockchain technology and cryptoassets in addressing financial inclusion. Koskelainen et al. (Reference Koskelainen, Kalmi, Scornavacca and Vartiainen2023) explored financial behavior in digital environments, concluding that digital financial literacy is distinct from both financial and digital literacy, and its measurement is relatively underdeveloped.
This paper contributes to the literature discussed above by measuring both crypto and financial literacy in Canada using the Bank of Canada’s Bitcoin Omnibus Survey (BTCOS). We utilize the crypto literacy measure proposed in Henry et al. (Reference Henry, Huynh and Nicholls2018a), which assesses understanding of three basic facts about the Bitcoin system and the Big Three financial literacy questions of Lusardi and Mitchell (Reference Balutel, Engert, Henry, Huynh and Voia2011a). We explore the intersection of these measures for two distinct cohorts – Bitcoin owners and non-owners – with a specific focus on the gender gap. Our analysis is primarily descriptive, with an emphasis on presenting relevant statistics to elucidate the current landscape of financial and crypto literacy in Canada. This builds on previous research (Balutel et al., Reference Balutel, Engert, Henry, Huynh and Voia2022a, Reference Balutel, Engert, Henry, Huynh and Voia2024) which documented much broader key findings from the BTCOS related to the awareness, ownership, and use of Bitcoin by Canadians; analysis of crypto and financial literacy was only tangential to these earlier works.
Most closely related to the current paper is Bannier et al. (Reference Bannier, Meyll, Röder and Walter2019), which documented a similar gender gap in Bitcoin literacy. Aside from studying the U.S. (versus Canadian) Bitcoin landscape, their analysis concerns the overall US population, while our paper compares the crypto and financial literacy of both the overall population and the subpopulation of Bitcoin owners. Additionally, their approach is different from ours. A key assumption they make is that financial literacy can help explain crypto literacy – in particular, that the gender gap in financial literacy explains a sizable portion of the gender gap in crypto literacy. By contrast, we use a joint bivariate model to examine the two measures of crypto and financial literacy (and their potential gender gaps). This model allows us to capture the reciprocal influence between these two variables, rather than treating them as separate and independent. Our model assumes that there are both observed factors (such as demographic characteristics) and, importantly, unobserved factors which may affect crypto and financial literacy simultaneously. Such unobserved factors could account for traits such as confidence in answering literacy questions, risk tolerance, or experience with conventional risky assets.
The key findings of our paper are as follows:
-
Crypto literacy. We find that women performed worse than men on the crypto literacy measure. Notably, the gender gap is present among Bitcoin owners: women who own Bitcoin indicated a lower understanding of the key features of the cryptoasset compared to their male counterparts. This could be related to our subsequent finding that women tended to choose “don’t know” responses to the crypto literacy questions more often than men, which is consistent with previous studies that have explored confidence as a possible source of the gender gap in literacy measures (Bucher-Koenen et al. Reference Bucher-Koenen, Lusardi and Alessie2017, Reference Bucher-Koenen, Alessie, Lusardi and Van Rooij2021; Cupák et al. Reference Cupák, Fessler, Hsu and Paradowski2020; Hospido et al. Reference Hospido, Iriberri and Machelett2024).
-
Financial literacy. We find no significant gender differences in financial literacy among Bitcoin owners. However, women performed worse than men on the financial literacy measure non-owners. Just under a third of women in the sample of non-owners are classified as having a high level of financial literacy, compared to approximately half of men. Given the low rate of Bitcoin ownership in Canada, this could indicate broader trends in financial literacy, consistent with the existing literature on the gender gap (Lusardi and Mitchell, Reference Lusardi and Mitchell2011a).
-
Joint crypto and financial literacy. The joint conditional analysis empirically supports the existence of a gender gap in both crypto and financial literacy among non-owners but only in crypto literacy among Bitcoin owners. Secondly, it reveals that crypto and financial literacy are not independent. There exists a positive and statistically significant correlation between the two scores, which can be attributed to unobservable factors (positive selection). This implies that individuals who demonstrate higher levels of crypto literacy are more likely to exhibit higher levels of financial literacy and vice versa. Moreover, this effect is more pronounced among Bitcoin owners in comparison to non-owners.
The rest of our paper is organized as follows: Section 2 describes the data used in the analysis and the construction of the literacy measures; Section 3 presents the unconditional analysis related to crypto and financial literacy among Bitcoin owners and non-owners with a focus on gender differences; then Section 4 discusses the conditional analysis of crypto and financial literacy. Section 5 concludes and suggests future work.
2. Bitcoin omnibus survey
This paper uses data on cryptoasset ownership from the Bank of Canada’s Bitcoin Omnibus Survey (BTCOS), specifically analyzing three waves conducted in 2018, 2019, and 2021. Footnote 7 The BTCOS is designed by the Bank of Canada, and fieldwork is conducted by the market research firm Ipsos. The survey uses an online and device-agnostic methodology, meaning that it can be completed on any device – computer, laptop, mobile phone, tablet, etc. – that can be used to access the internet. Quota-based sampling is used for recruiting respondents to the BTCOS to match nested population targets defined by age, gender, and region. The final sample sizes were 1,987 in 2018, 1,987 in 2019, and 1,974 in 2021. Among these, there are 99, 89, and 121 Bitcoin owners each year, respectively. The data are cleaned and the sample is weighted using an iterative raking procedure (Deville et al. Reference Deville, Sarndal and Sautory1993) to produce survey weights that ensure it is representative along numerous demographic dimensions as measured by the 2016 Canadian census. Footnote 8
2.1. Development of the BTCOS
The inaugural version of the BTCOS was conducted in 2016 as a pilot study, with a narrow focus on measuring levels of awareness and ownership of Bitcoin in Canada. With the exception of 2020, the survey has since been conducted annually during the month of December. Following the pilot, each subsequent version of the survey instrument contained additional content aimed at better understanding factors driving Bitcoin adoption. The BTCOS has served to inform the Bank of Canada’s view of future cash demand and to assess whether private digital currencies are making inroads as viable alternative payment methods. The latter forms one of the criteria for potential issuance of a Central Bank Digital Currency (CBDC), as outlined in Lane (Reference Lane2020). For previous versions of the BTCOS, see also the following reports: Henry et al. (Reference Henry, Huynh and Nicholls2017, Reference Henry, Huynh and Nicholls2018b, Reference Henry, Huynh, Nicholls and Nicholson2019b), and Balutel et al. (Reference Balutel, Felt, Nicholls and Voia2022b, Reference Balutel, Felt, Nicholls and Voia2023a).
The survey instrument covered various aspects related to Bitcoin, such as awareness, ownership, reasons for ownership/non-ownership, use in payments or person-to-person transactions, holdings, beliefs about its future survival and adoption levels among Canadians, price expectations over the next month, and methods of purchasing. The BTCOS survey also included inquiries about altcoins, alternative cryptocurrencies to Bitcoin, and addressed security incidents or price crashes experienced by both Bitcoin and altcoin owners. Additionally, for both cryptoasset owners and non-owners, the BTCOS gathered information about their cash holdings. And finally, of particular relevance to this paper, it assessed their levels of crypto and financial literacy. Footnote 9
2.2. Crypto and financial literacy measures
Exploring the intersection of finance and digitalization, three pivotal themes were outlined by Koskelainen et al. (Reference Koskelainen, Kalmi, Scornavacca and Vartiainen2023): fintech, financial behavior in digital environments, and behavioral interventions. A key observation from the study is that the measurement of digital financial literacy has remained underdeveloped compared to the measurement of financial or digital literacy.
The BTCOS used separate measures to assess crypto and financial literacy, shown in Tables 1 and 2. The crypto literacy questions were developed by Henry et al. (Reference Henry, Huynh and Nicholls2018a) and consist of three true-false statements concerning basic facts about the Bitcoin system. Bitcoin is the most well-known and market-dominant cryptoasset, and therefore knowledge of these facts reflects how well versed a person is in this domain. The questions are featured in a national library of financial literacy measures, Footnote 10 which supports the country’s primary consumer protection agency in its efforts to monitor and improve Canadians’ skills in navigating the financial marketplace (Financial Consumer Agency of Canada 2021). The measure has also been referenced in Bannier et al. (Reference Bannier, Meyll, Röder and Walter2019), and one of the questions used to assess digital financial literacy in OECD (2022) is similar to the true-false statement concerning government backing.
Crypto literacy questions
Note: This table shows the three questions Henry et al. (Reference Henry, Huynh and Nicholls2018a) used to test Bitcoin knowledge in the 2021 Bitcoin Omnibus Survey. Bitcoin is the most well-known and market-dominant cryptoasset; therefore, knowledge of these basic facts serves to measure crypto literacy. Respondents are asked to answer whether they think each statement is true or false; alternatively, they can answer “Don’t know.” Correct answers are shown in bold.
Financial literacy questions
Note: This table shows the three financial literacy questions that were asked in the 2021 Bitcoin Omnibus Survey. Questions are taken from the “Big Three” of Lusardi and Mitchell (Reference Lusardi and Mitchell2011a). Correct answers are in bold.
The Big Three questions developed by Lusardi and Mitchell (Reference Balutel, Engert, Henry, Huynh and Voia2011a) measure financial literacy by assessing understanding of compound interest, inflation, and risk diversification. These questions have been used in surveys across the world (Lusardi and Mitchell Reference Lusardi and Mitchell2023) and have become standard in the literature.
In our paper, we compute a crypto and financial literacy score for each respondent as the number of correct answers minus the number of incorrect answers (“don’t know” responses do not contribute to the score). Incorrect answers are deducted in order to penalize survey respondents who make guesses, potentially associated with their risk tolerance levels. Therefore, this score can take on integer values from
$ - 3$
to
$3$
. Crypto literacy is then classified as “high” (score
$=3$
), “medium” (score
$=1,2$
), or “low” (score ≤
$0$
). Financial literacy categories are constructed in the same way.
Our scoring methodology diverges from the conventional approach found in the literature (Lusardi and Mitchell Reference Lusardi and Mitchell2011c), which typically classifies individuals answering all three questions correctly as having high literacy and the rest as having low literacy. By assigning a value of zero to a “don’t know” answer and −1 to “incorrect” answers, our approach yields comparable insights to the conventional method for the high category, but distinguishes between two different types for those not in the high category. The decision to give a zero weight to a “don’t know” answer stems from the difficulty in distinguishing between a genuine lack of knowledge, a lack of confidence, or financial anxiety. Bucher-Koenen et al. (Reference Bucher-Koenen, Alessie, Lusardi and Van Rooij2021) identified that the use of “do not know” in response to financial knowledge questions by women is frequently linked to a lack of confidence, contributing to roughly one-third of the observed literacy gap, while Tinghög et al. (Reference Tinghög, Ahmed, Barrafrem, Lind, Skagerlund and Västfjäll2021) suggest financial anxiety stemming from a stereotype threat for women in the financial domain, which also plays a role in contributing to the observed gender gap.
3. Unconditional analysis of crypto and financial literacy
In this section, we provide a descriptive analysis of crypto and financial literacy levels among Canadians by aggregating data from 2018, 2019, and 2021 waves of the BTCOS survey. Our primary focus is on exploring the landscape of crypto and financial literacy among distinct ownership groups: non-owners (individuals without Bitcoin holdings) and Bitcoin owners, specifically emphasizing gender differences. Additionally, our analysis delves into how demographic factors influence crypto and financial literacy. Lastly, we present the distribution of combined crypto and financial literacy levels among Bitcoin owners and non-owners.
3.1. Crypto and financial literacy: Overall results
In general, awareness of the term “Bitcoin” among the Canadian population is high and has remained stable at about 90 percent since 2018 (Balutel et al., Reference Balutel, Engert, Henry, Huynh and Voia2022a, Reference Balutel, Engert, Henry, Huynh and Voia2024; Henry et al. Reference Henry, Huynh, Nicholls and Nicholson2020). However, the left panel of Figure 1 shows that the level of understanding of how Bitcoin actually works – i.e., the level of crypto literacy – is still quite low. Just 5 percent of non-owners displayed high crypto literacy, while a substantial 64 percent were identified with low literacy. Among Bitcoin owners, the distribution of crypto literacy reflects varying levels of understanding. Surprisingly, 32 percent exhibit a low understanding of Bitcoin features, while conversely, 31 percent have a high level of crypto literacy. The remaining 37 percent fall in the middle, demonstrating a moderate level of crypto literacy.
Crypto and financial literacy: Overall results.
Note: This figure shows the share of non-owners and Bitcoin owners in each category of crypto literacy (left) and financial literacy (right). The working sample size is 1,787 in 2018, 1,745 in 2019, and 1,778 in 2021. The sample comprises 99 Bitcoin owners in 2018, 89 in 2019, and 226 in 2021. All estimates are calculated using survey weights.
The right panel of Figure 1 shows the distribution of financial literacy for the pooled data. Specifically among non-owners, 39 percent have a high level of financial literacy, 35 percent are in the medium category, and 25 percent are in the low category. In contrast, a clear polarization in financial literacy is evident within the cohort of Bitcoin owners. Specifically, 40 percent of Bitcoin owners are categorized as having low financial literacy, while 37 percent fall into the high financial literacy category.
An observed trend among Bitcoin owners, as reported by Balutel et al. (Reference Balutel, Felt, Nicholls and Voia2023a) and Balutel et al. (Reference Balutel, Engert, Henry, Huynh and Voia2024), indicates a doubling percentage of individuals with low crypto literacy, climbing from 19 percent in 2018 to 40 percent in 2021. Conversely, the overall share of Bitcoin owners with high financial literacy has steadily increased. The price of Bitcoin soared during the COVID-19 pandemic, attracting investors looking for quick profits, but many jumped in without fully understanding the market complexities and risks involved. The 2021 BTCOS includes a novel question measuring the duration of Bitcoin ownership: “When did you first obtain Bitcoin?”. Footnote 11 Figure 2 shows that long-term owners, commonly known as early adopters, generally demonstrate better performance on crypto and financial literacy measures than recent owners (those who purchased Bitcoin starting in 2020), often referred to as late adopters. This difference could be attributed to the advantage early adopters have in terms of their more prolonged exposure and engagement in the market, while the entry of short-term owners may be characterized by hype and widespread media attention.
Crypto and financial literacy of Bitcoin owners: Long-term versus recent.
Note: This figure shows the share of long-term and recent Bitcoin owners in each category of crypto literacy (left) and financial literacy (right).Long-term owners, those who bought Bitcoin before 2020, and recent owners, who made purchases in 2020 or 2021, together form a sample of 226 (105 long-term, 121 recent). Data are from the Bank of Canada’s 2021 Bitcoin Omnibus Survey, specifically from the wave where the question “When did you first obtain Bitcoin?” was included.
Table 3 presents the proportions of crypto and financial literacy among Canadian Bitcoin owners according to various demographic categories. Footnote 12 When examining the results of crypto literacy by gender, it is observed that females have a higher proportion in the low crypto literacy category (43%) compared to males (28%). Conversely, males have a higher proportion in the high category (37%) compared to females (17%). Additionally, low crypto literacy is prevalent among Bitcoin owners with high school diplomas.
Demographics of Bitcoin owners by crypto and financial literacy
Note: This table reports the shares of Canadian Bitcoin owners according to their level of crypto and financial literacy. The sample consists of 414 Bitcoin owners. Years: 2018, 2019, and 2021. The Prairies region includes Alberta, Saskatchewan, and Manitoba. The Atlantic region includes New Brunswick, Nova Scotia, Prince Edward Island, and Newfoundland and Labrador. All estimates are calculated using survey weights.
The distribution of financial literacy tends to be split between the low and high categories across most demographic groups. When examining the results by gender, it shows that females have a slightly higher proportion of individuals with low financial literacy (43%) compared to males (38%). Nevertheless, both males and females exhibit a similar share in the high category (37%). When other demographic factors are examined, Bitcoin owners with a high school diploma or low income are more inclined to possess low financial literacy. In contrast, older owners with a university degree or high income tend to have higher financial literacy.
3.2. Crypto and financial literacy: Gender differences
This section explores unconditional gender-specific disparities in crypto and financial literacy within Bitcoin owner and non-owner cohorts. Revealing gender-specific patterns can help to ensure that various investors understand the risks they undertake when investing in cryptoassets.
Figure 3a presents crypto literacy distributions based on Bitcoin ownership and gender. Among non-owners, 56 percent of males and 71 percent of females fall into the low crypto literacy category, while 28 percent of males and 43 percent of females demonstrate low crypto literacy among Bitcoin owners. Additionally, a small percentage of non-owners (7% of males and 2% of females) exhibit high crypto literacy, while owners, particularly males (37%), surpass females (16%) in high crypto literacy. These results underscore the persistent gender gap in crypto literacy, with women consistently exhibiting a lower understanding of Bitcoin characteristics across both ownership categories. If financial literacy is associated with participation in asset markets, does this suggest that female Bitcoin owners have higher financial literacy levels than their female counterparts who do not own Bitcoin? If so, could this imply that the gender gap is narrower when comparing female Bitcoin owners with male Bitcoin owners?
Crypto and financial literacy: Gender differences.
Note: This figure shows the share of respondents in each category of crypto literacy (panel a) and financial literacy (panel b). The left side of each figure shows the distributions among non-owners of Bitcoin, categorized by gender, while the right side shows the distribution among Bitcoin owners, also categorized by gender. Categories are constructed based on scores described in Section 2.2. Years: 2018, 2019, and 2021. The sample size comprises 2,846 females, with 128 owning Bitcoin, and 2,464 males, with 286 being Bitcoin owners. All estimates are calculated using survey weights.
As illustrated in Figure 3b, the distribution of financial literacy reveals interesting insights regarding gender differences among Bitcoin owners and non-owners. Among non-owners, a clear gender gap is evident in the low financial literacy category, where 32 percent of females fall into this bracket, compared to 19 percent of males. In contrast, among Bitcoin owners, the gender difference in low financial literacy is less pronounced, with 43 percent of females and 38 percent of males falling into this category. In the medium financial literacy category, gender differences are less significant for both non-owners (39% females, 31% males) and owners (21% females, 25% males).
The high financial literacy category displays a noteworthy gender shift. Among non-owners, males dominate the high financial literacy category, constituting 50 percent compared to 29 percent of females. Surprisingly, among Bitcoin owners, the gender gap in high financial literacy narrows, with 37 percent of both males and females falling into this category. Upon an examination across the two groups of Bitcoin owners and non-owners, the most noticeable contrast emerges among males, where Bitcoin owners exhibit lower financial literacy compared to non-owners. Footnote 13 Nevertheless, among females, there are no notable average differences in financial literacy test scores, despite the fact that female Bitcoin owners possess a higher proportion of individuals with advanced financial literacy. The results suggest a nuanced relationship between gender and financial literacy, with significant disparities among non-owners and a more balanced distribution among Bitcoin owners.
Table 4 further decomposes the demographic patterns among female Bitcoin owners. Footnote 14 According to the findings, women with the lowest levels of understanding when it comes to cryptocurrency are typically found within two particular age ranges: 18–34 and over 55. They tend to have either a high school diploma or a lower income. However, females who fall within the 35–55 age range, possess a university degree, or earn between CAD 30k and 69k annually are more likely to exhibit a moderate level of crypto literacy. With regard to the financial literacy of female Bitcoin owners, those who fall within the low literacy category are typically aged 18–34, with a high school diploma or a low income. On the other hand, those who fall within the high literacy category are aged between 35 and 54, possess a university degree, or earn a high income. While the sample size is small, the results align with findings from other surveys on general financial literacy, which show that financial knowledge is lowest among younger age groups and that it is correlated with educational attainment (Lusardi and Mitchell Reference Lusardi and Mitchell2011c).
Demographics of female Bitcoin owners by crypto and financial literacy
Note: Share of female Bitcoin owners by crypto and financial literacy category. The sample size consists of 128 female Bitcoin owners. Years: 2018, 2019, and 2021. All estimates are calculated using survey weights.
Existing literature indicates that women are generally less inclined to hold risky assets and tend to be more financially risk-averse than men (Almenberg and Dreber Reference Almenberg and Dreber2015; Charness and Gneezy Reference Charness and Gneezy2012; Jianakoplos and Bernasek Reference Jianakoplos and Bernasek1998). In addition, Alonso et al. (Reference Alonso, Jorge-Vázquez, Rodrguez and Hernández2023) found, in the case of Spain, that reduced female participation in the crypto market is influenced by factors such as a lack of investment experience in traditional assets, a general deficiency in knowledge about cryptocurrencies, and a limited understanding of concepts like blockchain. Building on these insights, a recent survey report from the Ontario Securities Commission adds a layer to the story (Ontario Securities Commission 2022). Compared to non-owners, crypto owners (holding either cryptoassets, crypto funds, or both) were more likely to hold a variety of investments, particularly individually held stocks or exchange-traded units. Another noteworthy finding is that individuals acquiring cryptoassets primarily relied on their friends, family, and colleagues as a source of information before making purchases. Notably, this reliance on personal networks is more pronounced among females. Footnote 15 This interconnected narrative suggests that financially literate women may have acquired their knowledge and skills in more conventional financial markets, and a desire for portfolio diversification or exposure to alternative assets might drive their decision to invest in Bitcoin. However, the specific nuances of the cryptocurrency market, including its technology, decentralized nature, and unique risks, might not be as familiar to them.
3.3. Joint analysis of crypto and financial literacy
The observed disparities in crypto and financial literacy among Bitcoin owners require a thorough examination of their interaction. Table 5 presents valuable insights into the distribution of non-owners and Bitcoin owners based on their combined financial and crypto literacy levels.
Joint analysis of crypto and financial literacy: Non-owners and owners
Note: Share of non-owners and Bitcoin owners by crypto and financial literacy category. The sample size consists of 414 Bitcoin owners and 4,896 non-owners. Years: 2018, 2019, and 2021. All estimates are calculated using survey weights.
Non-owners aware of Bitcoin tend to have a significant percentage of people with low crypto literacy, regardless of their level of financial literacy. Specifically, 18 percent of individuals with low financial literacy, 24 percent with medium financial literacy, and 22 percent with high financial literacy have low crypto literacy. Moreover, only 3 percent of individuals have both high financial and crypto literacy.
Among Bitcoin owners, there is a diversity of literacy profiles. The share of owners with both low crypto and financial literacy (16%) is comparable to the share with high literacy across both dimensions (15%). The complexity of the relationship between financial and crypto literacy suggests that individuals may not uniformly apply their financial knowledge to the unique features of cryptocurrencies.
Table 6 presents the joint distribution of females based on their levels of both crypto and financial literacy and Bitcoin ownership. It indicates that females who do not own Bitcoin tend to have a limited understanding of cryptocurrencies, irrespective of their financial literacy. Among female Bitcoin owners, a larger percentage tends to fall into the low and medium categories of crypto literacy, irrespective of their financial literacy, with only 6 percent demonstrating high levels in both areas.
Joint analysis of crypto and financial literacy: female non-owners and owners
Note: Share of female non-owners and Bitcoin owners by crypto and financial literacy category. The sample size consists of 128 female Bitcoin owners and 836 female non-owners. Years: 2018, 2019, and 2021. All estimates are calculated using survey weights.
3.4. Gender differences in crypto and financial literacy: A decomposition based on literacy questions and response distribution
Bucher-Koenen et al. (Reference Bucher-Koenen, Lusardi and Alessie2017) have shown a significant gender gap in financial literacy worldwide, with women being more likely to answer “don’t know” to all financial literacy questions than men are. To gain deeper insights into this phenomenon, we conduct a detailed analysis of the response distribution of both male and female Bitcoin owners and non-owners for each crypto and literacy question. Footnote 16
3.4.1. Crypto literacy
Figure 4a compares the distribution of responses to each crypto literacy question. The percentage of correct answers is higher among owners than non-owners, irrespective of gender. Another finding is that the true/false statement “Bitcoin is backed by the government” has the highest prevalence of correct answers among both males and females, regardless of whether they own Bitcoin. On the other hand, the question about Bitcoin’s use of distributed ledger technology (DLT) has the lowest correct response rate, closely followed by the question related to the limited supply of Bitcoin.
Male Bitcoin owners particularly outperformed their non-owner counterparts in the question about Bitcoin’s use of distributed ledger technology (DLT), with owners answering correctly (61%) at over triple the rate (20%) of non-owners. Among women, Bitcoin owners performed best relative to non-owners in the question about the limited supply of Bitcoin, with female Bitcoin owners selecting the correct answer (42%) at nearly four times the rate of their non-owner counterparts (11%). These results confirm that Bitcoin owners are indeed more crypto-literate than non-owners, irrespective of gender.
Figure 4a also reveals a gender gap in crypto literacy. Men outperformed women in all questions and across both ownership categories.
Particularly in this regard, the questions related to the limited supply of Bitcoin and DLT produced the greatest gender difference in the percentage of correct answers for non-owners. Among non-owners, men answered correctly (20%) at over double the rate of women (11%). Among Bitcoin owners, the gender gap in response accuracy is generally consistent across the three questions, with men more likely to answer correctly compared to women. However, the most significant gender difference is observed in the questions related to Bitcoin’s supply, where men answered correctly (61%), a rate 1.5 times higher than for women (42%).
Given that men correctly answer the crypto literacy questions more often than women, regardless of whether they own Bitcoin, does this also translate to fewer incorrect answers? Figure 4a shows that among Bitcoin owners, women relative to men tend to have a relatively lower percentage of correct answers and a higher percentage of incorrect answers for two of the three questions, with the exception being the DLT-related question, where the share of incorrect answers is higher among males.
Does this mean that women are more inclined to make an assertive guess even when they are unsure of the correct answer? Although this may be the case among Bitcoin owners, among non-owners the proportion of incorrect answers by men exceeds that of women to all questions. This is most evident in the DLT question, with men answering incorrectly (27%) at nearly double the rate of women (14%).
Notably, we find that women chose “don’t know” more frequently than men in both ownership groups across all the three crypto-related questions. This is most evident in the distributions of answers to the DLT question showing that 32 percent of women selected “don’t know” compared with just 13 percent of men. This suggests that confidence in one’s answers may influence the measurement of crypto literacy, as is the case when measuring financial literacy (Bucher-Koenen et al. Reference Bucher-Koenen, Lusardi and Alessie2017). We formally explore the prevalence of “don’t know” responses in a forthcoming paper (Balutel et al., Reference Balutel, Henry, Huynh and Voia2023b).
3.4.2. Financial literacy
Figure 4b describes gender differences in the financial literacy of Bitcoin owners and non-owners by comparing the distribution of responses to each of the Big Three questions of Lusardi and Mitchell (Reference Lusardi and Mitchell2011a). The question related to interest compounding had the highest percentage of correct answers, while the question measuring knowledge of inflation had the largest share of incorrect answers, regardless of ownership status or gender.
Nevertheless, important gender differences are present. Among non-owners, women are less likely than men to answer financial literacy questions correctly. Similar to previous work that explored gender differences in financial literacy (Bucher-Koenen et al. Reference Bucher-Koenen, Lusardi and Alessie2017, Reference Bucher-Koenen, Alessie, Lusardi and Van Rooij2021), the question measuring knowledge of risk diversification proved to be especially challenging for women, with men choosing the correct answer (65%) at over 1.4 times the rate of women (48%).
Figure 4b suggests that female Bitcoin owners correctly answered the financial literacy questions in similar proportions to their male counterparts, except for the risk diversification question. Relative to the distribution of correct answers among Bitcoin owners, the gender differences are much more pronounced among non-owners.
Further, comparing the proportions of incorrect answers reinforces the notion that female Bitcoin investors may be just as financially literate as their male counterparts. Male owners answered incorrectly more frequently than women to one of the three questions. The only exception is the question pertaining to risk diversification and interest compounding, where the proportion of incorrect answers is comparable across genders.
Further, Figure 4b shows that the proportion of “don’t know” to the interest compounding and inflation questions among owners is comparable between men (5%, respectively 8%) and women (6%, respectively 9%). However, women were more likely than men to select “don’t know” for the risk diversification question. In contrast, female non-owners appear less confident in their knowledge across all three questions. Footnote 17
In sum, this analysis suggests that while the financial literacy gender gap is prevalent among non-owners of Bitcoin in Canada, female Bitcoin owners appear to be just as financially literate as their male counterparts. The lack of gender differences observed among crypto investors might also be attributed to a lower financial literacy score among male crypto owners when compared to their non-owner counterparts.
4. Conditional analysis of crypto and financial literacy
This section considers a comprehensive analysis aimed at unraveling the potential correlation between crypto and financial literacy. This investigation contributes to the discussion of Lyons and Kass-Hanna (Reference Lyons and Kass-Hanna2021), who underscored the nascent nature of research that defines and measures digital literacy while establishing connections to financial literacy to understand the pathways to financial behavior. Kass-Hanna et al. (Reference Kass-Hanna, Lyons and Liu2022) also delved into the combined impacts of financial and digital literacy. They constructed a composite index for digital and financial literacy (DFL), revealing that both financial and digital literacy are pivotal factors in shaping positive financial behaviors and ensuring long-term financial security.
Unlike Bannier et al. (Reference Bannier, Meyll, Röder and Walter2019), which documented a gender gap in Bitcoin literacy among the U.S. population assuming that financial literacy can help explain crypto literacy, we examine the interdependencies between crypto and financial literacy of Bitcoin owners and non-owners. Footnote 18 In light of a plausible bidirectional relationship between crypto and financial literacy (potential for simultaneous learning in both crypto and financial), we employ a bivariate ordered probit model (Greene and Hensher Reference Greene and Hensher2009; Sajaia Reference Sajaia2008). Footnote 19
Our model assumes that there are both observed factors (such as demographic characteristics) and, importantly, unobserved factors which may affect crypto and financial literacy simultaneously. Such unobserved factors could account for traits such as confidence in answering literacy questions, risk tolerance, or experience with conventional risky assets.
Table 7 provides the joint estimation results for the crypto and financial literacy scores for Bitcoin owners (Columns 2–3) and non-owners (Columns 4–5), accounting for demographic characteristics, province fixed effects, and time effects. In the realm of crypto literacy, being female among both Bitcoin owner and non-owner cohorts has a negative and statistically significant effect. This suggests that women, irrespective of ownership, are less likely than men to possess knowledge about Bitcoin.
Conditional analysis of crypto and financial literacy
Note: Estimates of crypto and financial literacy scores (low, medium, and high) using a joint order probability model (bivariate ordered probit) for Bitcoin owners subsample (Columns 2–3) and non-owners subsample (Columns 4–5). The base categories are male, aged 18 to 34 years, with high school education, low income (less than $30,000 per year), from British Columbia, and unemployed. Years: 2018, 2019, and 2021. Athrho represents the correlation for the two errors in the bivariate ordered probit model. All estimates are calculated using survey weights.
Shifting our focus to financial literacy, our results indicate that being female is negative and statistically significant, particularly among non-owners. However, these coefficients are not statistically significant for the Bitcoin owners subsample. This indicates that women’s financial literacy tends to lag behind men’s among non-owners. However, there is no evidence suggesting a significant gender-based difference in financial literacy between female and male Bitcoin owners.
Further, we discuss other demographic characteristics that impact crypto and financial literacy of Bitcoin owners and non-owners. In this regard, age does not consistently impact crypto and financial literacy among owners and non-owners. Bitcoin owners aged 35–54 and above 55 years old tend to have higher levels of financial literacy and crypto literacy, although the results show statistical significance only for the financial literacy of those over 55. Non-owners in both of these age groups tend to have lower crypto literacy but a stronger understanding of financial literacy. It appears that older individuals have less knowledge about the features of cryptoassets compared to younger people. Nevertheless, if they do own cryptoassets, they usually exhibit better financial literacy than their younger counterparts. Attaining a university degree positively affects both crypto and financial literacy, irrespective of ownership status. However, while possessing a university degree positively impacts crypto literacy for Bitcoin owners, this impact is not statistically significant. High-income levels (>70K) do not affect crypto literacy but positively affect financial literacy regardless of ownership status.
Furthermore, we document that unobserved factors may simultaneously account for both crypto and financial literacy, with a more pronounced impact on the cohort of Bitcoin owners. Footnote 20 For example, if something drives an increase in financial literacy, it may also drive an increase in crypto literacy and vice versa. Crypto owners, typically young males influenced by FOMO or peer pressure, may initially have lower financial literacy. However, investing in crypto could improve their literacy and spark an interest in the stock market, further improving their financial literacy. Conversely, individuals with high financial literacy and stock market experience might want to diversify their portfolio by exploring cryptoassets, thereby enhancing their crypto literacy.
This exploratory work can form the basis of future empirical studies. In particular, accounting for selection into Bitcoin ownership would provide a relatively more accurate comparison of the differences in crypto and financial literacy between genders among cryptoasset owners – i.e., a more “apples-to-apples” comparison. This could be accomplished by matching the two samples of Bitcoin owners and non-owners using program evaluation techniques such as propensity score matching, inverse probability weighting, or regression adjustment, among others.
5. Conclusion and discussion
In this paper, we used survey data from a nationally representative sample of Canadians to reconcile low female participation in the cryptoasset market with the well-documented gender gap in financial literacy. Cryptoassets are complex products primarily used as investments, but their key features are sufficiently distinct from conventional assets that making informed decisions could require knowledge of concepts not typically captured by financial literacy measures (Bannier et al. Reference Bannier, Meyll, Röder and Walter2019). We use a novel measure of crypto literacy in conjunction with a measure of financial literacy to assess this perspective.
We find that Canadian Bitcoin owners are more informed than non-owners about the asset and thus have higher levels of crypto literacy. They are heterogeneous with respect to financial literacy. We find that women who own Bitcoin are just as financially literate as their male counterparts but lag on crypto literacy. Moreover, the crypto literacy gender gap persists regardless of ownership status – women are generally less informed about cryptoassets.
These results indicate that measuring crypto literacy can usefully complement measures of financial literacy in the context of digital asset markets with complex investment products that may be less reliant on knowledge of traditional financial concepts. Better crypto knowledge and financial literacy, in turn, could help investors make better decisions and perhaps reduce the prevalence of FoMO as an investment motivation and might help decrease exposures in such markets.
The results for crypto and financial literacy discussed in this paper might suggest wider implications. For instance, in order for consumers to make informed decisions with regard to initiatives like open banking and new developments in mainstream financial technology more generally, an understanding of such products is required. However, according to a recent press release from the Financial Consumer Agency of Canada, only 9 percent of survey respondents had heard of open banking, and once this term was explained to them, only 15 percent said they would participate in open banking initiatives (Financial Consumer Agency of Canada 2023). At the same time, only 18 percent understood the consumer protection provisions for services offered by FinTechs.
Similarly, a recent report on payment preferences commissioned by the European Central Bank (ECB) revealed that some individuals have reservations about using banking apps and digital payments because of their limited technical knowledge and skills, which led to fears about making errors during transactions (European Central Bank 2022). Related considerations probably would apply to the potential issuance of a CBDC. In this regard, a recent public consultation by the ECB found that (European Central Bank 2022): “Among the general public and the tech-savvy, there was little awareness of the digital euro. As a result, participants expressed a need for more information about “why” it is needed, and “how” it differs from the euro kept in bank accounts and spent electronically via apps or bank transfers…. [P]articipants associated digital euro with a cryptocurrency, and/or believed this was a digital money designed to replace cash…. Some described it as a blockchain-based technology that would replace money” (pp. 51–52).
In sum, a key issue that probably deserves more attention concerns the significant scale of investment in public education that would be required to inform consumers about such a product. Footnote 21
Acknowledgements
The views expressed in this article are those of the authors. The Bank of Canada is not responsible for any views expressed in the paper. All remaining errors are the authors’ responsibility. We thank Alexis Direr, David Fernandes, Annalisa Ferrando, Joann Jasiak, Henry Kim, Annamaria Lusardi, Francisco Rivadeneyra, and two anonymous reviewers for their valuable feedback. We thank Mario Caceres, Maryam Akbar, and team at Ipsos for improving and implementing the Bitcoin Omnibus Survey. We also thank participants of the Asia Pacific Financial Education Institute (APFEI), Canadian Economic Association, Bank of Finland Financial Literacy Conference, University of Urbino Carlo Bo Financial Literacy Conference, Bank of Canada’s Central Bank Currency Department and Digital Currency Econ Working Group, and participants of the York University Department of Economics and Laboratoire d’Économie d’Orléans, Université d’Orléans Seminar Series. Thank Maren Hansen for editorial support.
Appendix
Appendix A. 2021 BTCOS instrument
The 2021 BTCOS was completed by respondents entirely online through the web or on mobile devices. Below is a representation of the online survey instrument. Skip logic and other programming instructions are included between square brackets but were not shown to participants. Note that demographic questions and questions related to survey recruitment were also asked but are not shown here. Please cite Balutel et al. (Reference Balutel, Engert, Henry, Huynh and Voia2022a) if you wish to use these questions.
2021 BTCOS instrument.
Open access
