Group norms and policy norms trigger different autonomous motivations for Chinese investors in cryptocurrency investment

The paper investigates why, despite tightening regulations, Chinese investors continue to engage in cryptocurrency investment. Building on social norms theory, the authors posit that descriptive (group) norms and injunctive (policy) norms exert opposing influences on investors’ autonomous motivation to invest. Group norms may foster belongingness and positive modeling effects, thereby triggering autonomous motivation and increasing investment. Policy norms, by emphasizing sanctions and risks, heighten perceived risk and self-control, suppress autonomous motivation, and reduce investment. The study tests two core questions: the mechanism through which social norms influence cryptocurrency investment (via autonomous motivation), and boundary conditions (cryptocurrency knowledge) that change these effects. China’s strong regulatory stance coupled with active market participation provides a suitable context to test these hypotheses.

The literature identifies multiple psychological and social drivers of cryptocurrency investment, including sentiment, herding, social influence, financial knowledge, perceived behavioral control, trust, and suspicion. However, the policy environment’s role is underexplored. Drawing on social norms theory, descriptive norms (group norms) reflect perceptions of what others do, while injunctive norms (policy norms) reflect perceived approvals and prohibitions. Prior work shows social influence and peer effects can increase crypto adoption, whereas regulatory interventions can depress market activity and returns. The authors hypothesize: H1a: Group norms promote cryptocurrency investment; H1b: Policy norms inhibit cryptocurrency investment. Based on self-determination theory, autonomous motivation is proposed as a mediator (H2). Regarding knowledge as a boundary condition, the authors distinguish subjective knowledge (SK) and objective knowledge (OK). They hypothesize SK positively moderates the group norms–autonomous motivation link and negatively moderates the policy norms–autonomous motivation link, whereas OK negatively moderates the group norms–autonomous motivation link and positively moderates the policy norms–autonomous motivation link (H3a, H3b).

Design: Cross-sectional online survey of Chinese users attentive to cryptocurrency investment. Instruments were developed in English, translated and back-translated to Chinese; pretested with 256 responses, then refined. Data collection occurred June–July 2021 via an online platform. Sampling and screening: 1,100 total responses from 31 provinces/municipalities/autonomous regions. Validity checks included awareness screening items, logic traps, and repetitive-option detection, leading to exclusion of 353 responses (281 unaware/invalid, 72 failed attention/logic), yielding 727 valid cases (66% effective rate). Sample characteristics: 44.8% had prior crypto investment experience; 48.3% male; age and education broadly distributed (e.g., 24.9% aged 31–40; 27.4% aged 41–50; 35.5% undergraduate, 23.0% master’s). Investment experience and expenditure varied considerably. Measures: Seven-point Likert scales (1=strongly disagree; 7=strongly agree) used for multi-item constructs: Group norms (4 items; Ryu & Ko, 2019), Policy norms (4 items; adapted from Xie, 2019), Autonomous motivation (4 items; Mustafa & Ali, 2019; Kuvaas, 2006), Subjective knowledge (4 items; Ryu & Ko, 2019), Cryptocurrency investment intention (3 items; Palamida et al., 2018). Objective knowledge measured via 10 true/false/don’t know items about cryptocurrency (scored, then dichotomized: high OK ≥5 correct; low OK <5). Control variables: gender, age, education, investment experience, investment expenditure. Validation: Descriptive statistics and correlations computed. Reliability and validity supported: Cronbach’s alpha >0.90; KMO >0.7; standardized factor loadings >0.8; composite reliability >0.9; AVE >0.7; discriminant validity via square root of AVE exceeding inter-construct correlations. CFA model fit acceptable: χ2=471.08, χ2/df=3.32, GFI=0.93, CFI=0.98, IFI=0.98, RMSEA=0.06. Analysis: Multiple linear regression tested main effects (group norms, policy norms) on investment intention, controlling for demographics and investment-related controls. Mediation tested via bootstrapped PROCESS analysis (5,000 resamples, 95% CI) for autonomous motivation as mediator between norms and investment. Moderation tested for SK (continuous, mean-centered) and OK (binary high/low) on the relations between norms (group, policy) and autonomous motivation, with simple slopes analyses.

• Main effects (Table 4): Group norms positively predict cryptocurrency investment (standardized β=0.354, p<0.001). Policy norms negatively predict cryptocurrency investment (standardized β=−0.120, p=0.001). Model fit: F(7,719)=25.172, p<0.001; R2=0.197. H1a and H1b supported.
• Mediation (Fig. 2; Table 5): • Group norms → Autonomous motivation: β=0.416, p<0.001. Direct effect on investment: B=0.253, 95% CI [0.177, 0.329]; Indirect effect via autonomous motivation: B=0.139, 95% CI [0.097, 0.184]. Partial mediation. • Policy norms → Autonomous motivation: β=−0.417, p<0.001. Direct effect on investment: B=0.033, 95% CI [−0.046, 0.112] (ns); Indirect effect via autonomous motivation: B=−0.186, 95% CI [−0.232, −0.140]. Full mediation. H2 supported.
• Moderation by subjective knowledge (SK) (Table 6, Models 1–2): • Group norms → Autonomous motivation: main β=0.429, p<0.001; interaction GN×SK β=−0.174, p<0.001. SK weakens the positive effect of group norms. • Policy norms → Autonomous motivation: main β=−0.356, p<0.001; interaction PN×SK β=−0.145, p<0.001. SK strengthens the negative effect of policy norms. H3a not supported (direction opposite to prediction).
• Moderation by objective knowledge (OK) (Table 6, Models 3–4): • Group norms → Autonomous motivation: main β=0.517, p<0.001; interaction GN×OK β=−0.269, p<0.001. OK weakens the positive effect of group norms. • Policy norms → Autonomous motivation: main β=−0.231, p<0.001; interaction PN×OK β=−0.235, p<0.001. OK strengthens the negative effect of policy norms. H3b supported. Overall: Group norms increase, and policy norms decrease, crypto investment, with autonomous motivation as a key pathway. Both subjective and objective cryptocurrency knowledge consistently suppress autonomous motivation in these pathways.

The findings clarify how social norms affect cryptocurrency investment through autonomous motivation. Group norms cultivate belonging and positive modeling, elevating autonomous motivation and, in turn, investment intentions. Conversely, policy norms signal regulatory risk and potential penalties, reducing autonomous motivation and investment. Autonomous motivation fully accounts for the effect of policy norms on investment and partially for group norms, underscoring motivation as a central mechanism. Cryptocurrency knowledge serves as a boundary condition: higher subjective or objective knowledge dampens autonomous motivation arising from group norms and amplifies the suppressive influence of policy norms, indicating that informed investors exercise more self-control and are less swayed by group enthusiasm. These results integrate social norms theory with self-determination theory within a politically salient context, demonstrating that macro-level policy environments and micro-level knowledge jointly shape investor motivation and behavior. Practically, the study suggests that regulatory clarity and enforcement can effectively deter speculative participation by reducing autonomous motivation, while unchecked group-driven discourse may fuel speculative activity. Managing public information environments and improving investor knowledge can shift motivations toward caution.

This study contributes by integrating social norms (group and policy) with self-determination theory to explain cryptocurrency investment behavior in China. It shows that autonomous motivation mediates the effects of norms, with group norms increasing and policy norms decreasing investment via motivation. It also establishes cryptocurrency knowledge as a boundary condition that suppresses autonomous motivation across both pathways. The work extends self-determination theory to consumer investment contexts, particularly cryptocurrencies, and highlights the importance of the policy environment. Policy implications include reinforcing and clarifying crypto-related regulations and actively managing public discourse to curb speculative contagion, alongside education initiatives to raise both subjective and objective cryptocurrency knowledge. Future research should address identified limitations by employing objective behavioral data (e.g., transaction-level big data), improving and validating measurement of cryptocurrency investment, examining specific cryptocurrencies and their heterogeneity, expanding to multi-country and longitudinal designs to capture temporal and contextual dynamics, and distinguishing cryptocurrency-specific from general financial literacy effects.

• Self-reported survey data may suffer from response biases (over/under-reporting), affecting construct estimates.
• Measurement of cryptocurrency investment is still nascent and may lack full validation.
• The survey treated the top ten cryptocurrencies as representative of all cryptocurrencies; results may not generalize to all types or capture currency-specific effects.
• Use of a unified “cryptocurrency” concept may overlook idiosyncrasies of particular assets (e.g., Bitcoin vs. Dogecoin).
• Cross-sectional design from a specific time window (June–July 2021) and country (China) limits generalizability; results should be interpreted cautiously.
• Need for longitudinal and cross-country analyses to capture evolving regulation, markets, and investor behavior.