Most crypto studies focus on ownership. This risks missing important differences among non-owners: those who would not consider owning vs. those who would consider but haven't acted yet.
Cryptocurrency is a rapidly expanding, multi-trillion dollar landscape spanning sociopolitical, psychological, technological, and financial domains. Understanding who considers and owns crypto matters for communicators, policymakers, educators, and researchers interested in studying and/or reaching these groups.
Most research focuses only on owners vs. non-owners. Prior work that does account for considerers tends to use methods that don't allow for the direct comparison between rejecting, considering, and owning groups. As a result, we know little about broader patterns of individual-level crypto adoption behavior in the United States.
According to the Theory of Planned Behavior (Ajzen 1985 ↗; 2020 ↗; Fishbein and Ajzen 1975 ↗), people must think about and form intentions to engage in a behavior before taking action. However, positive considerations about a behavior don't always lead to intention, and intentions don't necessarily guarantee follow through.
We argue that considerers and owners may both differ from rejectors — but considerers aren't necessarily like owners in all ways or to the same degree.
—
TOTAL MARKET CAP
—
NO. OF CRYPTOCURRENCIES
Total crypto market capitalization and number of active cryptocurrencies.
We asked respondents if they owned crypto and, if not, whether they would consider it. In all three studies, ~20% would consider owning — suggesting considerers are a real, measurable group.
// REJECTORS
Does not own and would not consider owning.
SAMPLE RATES
68%
STUDY 1
62%
STUDY 2
46%
STUDY 3
// CONSIDERERS
Does not own but would consider owning.
SAMPLE RATES
19%
STUDY 1
20%
STUDY 2
22%
STUDY 3
// OWNERS
Current holders of cryptocurrency.
SAMPLE RATES
13%
STUDY 1
18%
STUDY 2
32%
STUDY 3
HOW MANY TYPES: CONSIDER OR OWN?
We then asked them to select the type(s) of cryptocurrency they would consider or owned. While most considerers and owners selected one type of cryptocurrency, many selected two or more types.
// WOULD CONSIDER
1 TYPE2+ TYPES
65% / 35%
STUDY 1
76% / 24%
STUDY 2
67% / 33%
STUDY 3
// OWNS
1 TYPE2+ TYPES
65% / 35%
STUDY 1
81% / 19%
STUDY 2
51% / 49%
STUDY 3
WHICH TYPES: CONSIDER OR OWN?
As far as which type(s) they would consider or owned, Bitcoin dominates — followed by Ethereum and other types (Dogecoin, XRP, etc.).
// WOULD CONSIDER
₿ BITCOIN
93%
STUDY 1
89%
STUDY 2
93%
STUDY 3
Ξ ETHEREUM
37%
STUDY 1
31%
STUDY 2
37%
STUDY 3
◈ OTHER
8%
STUDY 1
5%
STUDY 2
4%
STUDY 3
// OWNS
₿ BITCOIN
70%
STUDY 1
76%
STUDY 2
87%
STUDY 3
Ξ ETHEREUM
34%
STUDY 1
33%
STUDY 2
53%
STUDY 3
◈ OTHER
29%
STUDY 1
10%
STUDY 2
9%
STUDY 3
SAMPLE RATES MIRROR MARKET SHARES
Our sample rates are similar to previous research and real-world market shares, lending external validity to our measures and results.
58.2%
TOTAL MARKET SHARE OF BITCOIN
10.2%
TOTAL MARKET SHARE OF ETHEREUM
31.6%
TOTAL MARKET SHARE OF OTHER CRYPTO
Market share of Bitcoin relative to Ethereum and "other" cryptocurrencies.
We use a modeling approach that simultaneously estimates the probability of being a rejector, considerer, or owner. We repeat our analysis using three independent samples (different people, timing, and survey providers) to ensure the reliability of our results.
Multivariate, generalized ordered logistic regression (partial proportional odds) models — run separately for each of three independent, nationally-representative samples of US adults.
A three-level outcome capturing each crypto adoption stage: rejectors (does not own and would not consider), considerers (does not own but would consider), and owners (currently owns cryptocurrency). Each respondent is classified into exactly one group.
SOCIODEMOGRAPHICS— who respondents are
AGESEXRACE/ETHNICITYEDUCATIONINCOMERELIGIOSITY
PERSONALITY— how one thinks, feels, and approaches the world
OPENNESSCONSCIENTIOUSNESSEXTRAVERSIONAGREEABLENESSNEUROTICISMNEED FOR COGNITION
POLITICAL ORIENTATION— policy preferences
OPERATIONAL IDEOLOGY
PERSONALITY ADAPTATIONS— orientation toward authority and social order
NEED FOR CHAOSAUTHORITARIANISM
TRUST— confidence in institutions and others
GOVERNMENTPEOPLEBANKS
TECH OPTIMISM— attitudes toward emerging tech
SUPPORT FOR DEVELOPING AI
INVESTMENTS & RISK— financial behavior and risk appetite
STOCK OWNERINVESTING RISK TOLERANCE
Study 1 — Cooperative Election Study (Oct–Dec 2024, n = 803) Study 2 — Lucid (Feb–Mar 2025, n = 1,908) Study 3 — Verasight (June 2025, n = 2,086)
WHAT DO WE FIND?
IN ALL THREE SAMPLES, CONSIDERERS AND OWNERS ≠ REJECTORS
BUT CONSIDERERS ≠ OWNERS
— not in all ways, or to the same degree
FINDINGS I
Considerers and owners both differ from rejectors, and share similar demographics and personality traits.
Click a variable to see how it shifts the average probability of being a rejector, considerer, or owner (from min. to max. values, all else equal).
FINDINGS II
At the same time, some characteristics distinguish rejectors from considerers, while others distinguish considerers from owners.
Click a variable to see how it shifts the average probability of being a rejector, considerer, or owner (from min. to max. values, all else equal).
WHAT OUR FINDINGS MEAN FOR...
Understanding who considers, and what moves them from intentions to action, has direct implications for product design, onboarding, and communication strategies. Whether one's goal is to endorse or denounce crypto, the "considerer" is arguably the most actionable audience: potentially open to adoption, but hasn't taken action yet.
Our work suggests that no single approach will appeal to all. For example, efforts to reduce uncertainty and promote legitimacy may attract considerers, but backfire with owners who think differently about institutional disruption and risk.
That roughly 20% of US adults would consider owning cryptocurrencies suggests significant latent demand. However, consideration may depend on pending regulation, perceived risk, and other legitimacy signals, including endorsement from trusted elites. Policymakers in this area will likely need to balance both mainstream and alternative frames.
Looking only at ownership obscures variation among non-owners. By accounting for considerers, we find certain factors (e.g., ideology and tech attitudes) do matter for crypto adoption–just not for the ownership stage. Our approach also helps to determine when other factors (e.g., Need for Chaos) matter in the crypto adoption process, providing important direction for future research.
Considering and owning rates from all three studies suggest US crypto adoption is at or nearing the "early majority" stage. But we don't yet know how the crypto landscape will evolve, or whether the motivations of the next "wave" of adopters will look the same. Regardless, our findings align with decades of attitude-behavior research and should be broadly applicable even amid variation in how cryptocurrency is understood and/or used.
Most respondents in our samples would consider or own Bitcoin, but a sizable share have multiple types and/or don't engage with Bitcoin at all. While it might be tempting to focus exclusively on a single, high-profile asset like Bitcoin, doing so presents several methodological problems.
First, limiting the analysis to Bitcoin ignores a segment of the otherwise crypto-engaged. For those who hold Bitcoin alongside other cryptocurrencies, this approach makes it virtually impossible to disentangle whether any observed relationships are unique to Bitcoin itself. For example, if younger individuals are more likely to own Bitcoin and other crypto, we can't conclude that age predicts Bitcoin ownership per se.
Focusing on a single asset also overlooks the fact that, for many, terms like "Bitcoin" and "crypto" are synonymous. Understanding whether certain factors are indeed asset-specific requires accounting for the myriad cryptocurrencies, then empirically testing whether differences between them exist.
We tested for, but did not find, asset-specific differences in our data. In other words, it's possible for such asset-specific differences to emerge, but our multi-study evidence suggests this is not currently the case.