Do psychological costs matter? The mechanism of perceived value on green consumption behaviour

The study addresses why consumers engage in green consumption beyond traditional explanations by incorporating psychological costs into the perceived value–behaviour framework. Motivated by SDG 12 and the need to expand green behaviours beyond purchase to recycling and refusal of disposables, the authors use mindsponge theory and diminishing marginal utility to examine how perceived hedonic and utilitarian values influence green consumption via willingness to pay a green premium (WPGP), and how psychological costs (stress, stigma, autonomy) shape this mechanism. Research questions: (1) How do perceived values correlate with green consumption through WPGP? (2) Do hedonic and utilitarian values differ in their effects on green consumption via WPGP? (3) How do psychological costs influence the perceived value–WPGP–green behaviour mechanism? The study aims to inform nuanced, psychologically informed policy interventions for promoting sustainable consumption in China.

Green consumption encompasses purchasing green products, recycling, and refusing disposable products. Prior work explains green behaviours through individual factors, product/service attributes, and macro-environment/policy. Perceived value, often multidimensional, is distilled here into utilitarian (functional/economic) and hedonic (affective/pleasure) dimensions. Many studies show perceived value promotes green purchase, recycling, and repurchase intentions, though findings are not always consistent. WPGP, reflecting extra monetary/time/effort costs of green options, is often treated as an intention proxy but evidence on its link to behaviour is mixed. The authors posit WPGP mediates effects of perceived values on green behaviours (H1) and that hedonic value may more strongly drive WPGP (and thus behaviours) than utilitarian value when a premium is involved (H2), while utilitarian value may exert stronger direct (non-WPGP) effects (H3). Psychological costs—stress (from expense/effort and pressure), stigma (felt if not acting green), and autonomy (perceived loss of choice)—may directly affect WPGP (H4a–H4c) and moderate perceived value→WPGP links (H5a–H5c), with differing mechanisms per mindsponge theory and diminishing marginal utility: stress/autonomy raise costs of engaging in green, potentially heightening sensitivity to perceived value; stigma represents costs of not engaging in green, potentially dampening marginal utility of perceived value when high.

Design and sample: Cross-sectional online survey on the Credemo platform (China) from March 23 to April 2, 2023; N=677; participants received 5 RMB. The sample skewed younger, more educated, higher income, and more female than the general population; diverse employment statuses. Covariates included age, sex, education, income, and work. Procedure: After consent and a brief definition of green consumption, respondents completed randomized scales (~5–10 minutes). Measures: Latent constructs included perceived hedonic value (PHV), perceived utilitarian value (PUV), WPGP, three green behaviours (purchase green goods, recycle unused items, refuse/use disposable products), and psychological costs (stress, stigma, autonomy). Items are provided in Supplementary Appendix A (Table A1). Analysis: Conducted in R 4.3.0. Confirmatory factor analysis (CFA) using lavaan established the measurement model for nine main variables; factor scores and correlations computed. Model fit for the structural measurement model was good (χ²=460.874, df=216, CFI=0.961, GFI=0.944, RMSEA=0.041, SRMR=0.044). Reliability/validity: Cronbach’s α and Composite Reliability (CR) mostly >0.7 (autonomy ≈0.63–0.65 acceptable); AVE mostly >0.5 with acceptable convergent validity; Fornell-Larcker criteria supported discriminant validity. Common method bias assessed via Harman’s single-factor test: first factor explained 27.6% (<50%). Hypothesis testing: Structural equation modeling (SEM) with moderated mediation path model (per Fig. 1) using lavaan; 5000 bootstrap resamples produced bias-corrected 95% CIs for indirect and conditional indirect effects. Control variables were included on mediator and outcomes.

Direct effects (excluding WPGP): Utilitarian value significantly increased purchase of green goods and recycling, and reduced use of disposable products; hedonic value significantly increased purchase only. Differences: Utilitarian value’s negative effect on disposable use stronger than hedonic (ΔPHV−PUV=0.352, 95% CI [0.201, 0.493]); utilitarian’s positive effect on recycling larger than hedonic (ΔPHV−PUV=−0.172, 95% CI [−0.304, −0.045]); no significant difference for green purchase (Δ=−0.077, 95% CI [−0.219, 0.059]). Mediation via WPGP (Table 3): WPGP significantly mediated perceived values→behaviours. Indirect effects (standardized): PHV→WPGP→Use disposable products: −0.043 (p=0.011; 95% CI −0.079, −0.013); PHV→WPGP→Purchase green goods: 0.073 (p<0.001; 95% CI 0.042, 0.110); PHV→WPGP→Recycle: 0.069 (p<0.001; 95% CI 0.037, 0.108). PUV equivalents: −0.023 (p=0.012; 95% CI −0.043, −0.007); 0.040 (p=0.002; 95% CI 0.017, 0.067); 0.038 (p=0.003; 95% CI 0.016, 0.065). Hedonic indirect effects exceeded utilitarian for purchase (Δ=0.033, 95% CI 0.003, 0.068) and recycling (Δ=0.032, 95% CI 0.003, 0.067); small difference for disposables (Δ=−0.019, 95% CI −0.050, −0.001). Psychological costs and WPGP: Stress decreased WPGP (p=0.004); stigma increased WPGP (p<0.001); autonomy had no significant association (p=0.549). Moderation on value→WPGP: PHV×Stress positive (β=0.112, p=0.043), strengthening PHV→WPGP; PUV×Stigma negative (β=−0.137, p<0.001), weakening PUV→WPGP. PHV×Stigma, PUV×Stress, and all autonomy interactions were non-significant. Moderated mediation (Table 4): Under high stress, PHV indirect effects strengthened: to disposables −0.058 (p=0.013), purchase 0.099 (p<0.001), recycling 0.095 (p<0.001); under low stress they were smaller (e.g., purchase 0.046, p=0.022). Under low stigma, PUV indirect effects were significant (e.g., purchase 0.072, p=0.002; recycling 0.069, p=0.001; disposables −0.042, p=0.009); under high stigma, they attenuated to near-zero and non-significant. Comparative conditional indirects (Table 5): PHV indirect effects exceeded PUV under high stress (e.g., purchase Δ=0.063, 95% CI 0.008, 0.135; recycling Δ=0.060, 95% CI 0.007, 0.134) and under high stigma (e.g., purchase Δ=0.077, 95% CI 0.033, 0.129; recycling Δ=0.074, 95% CI 0.031, 0.124). Autonomy did not significantly moderate, though at low autonomy PHV’s indirects tended to exceed PUV’s.

Findings clarify the mechanism linking perceived value to green behaviours. WPGP is a significant mediator: higher perceived value raises willingness to bear the green premium, which increases purchasing green, recycling, and refusing disposables. Hedonic value outperforms utilitarian value in driving WPGP and thus behaviours when a premium is involved, consistent with mindsponge theory and marginal utility logic: paying extra clashes with utilitarian cost-effectiveness, whereas hedonic benefits can justify premiums. Directly, utilitarian value better promotes recycling and refusal of disposables when WPGP is excluded. Psychological costs matter: stress both depresses WPGP and intensifies the hedonic value→WPGP link, making hedonic appeals more potent under high stress; stigma elevates WPGP but diminishes the reliance on utilitarian value (alternative benefit of stigma reduction reduces marginal utility of utilitarian gains). Autonomy showed no direct or moderating influence. Policy relevance: when green actions entail premiums or effort, emphasizing hedonic/affective benefits is more effective, especially for consumers experiencing higher psychological costs; where increasing WPGP is infeasible, highlighting utilitarian/functional value can still encourage behaviours that do not require premiums.

An online survey of 677 Chinese consumers using SEM shows that perceived values influence green consumption both directly and indirectly via WPGP. Hedonic and utilitarian values play distinct roles: hedonic value more strongly promotes green behaviours through WPGP when premiums are involved, while utilitarian value exerts stronger direct effects on recycling and reducing disposable use. Psychological costs shape these mechanisms: stress strengthens the hedonic value→WPGP path and reduces WPGP overall; stigma increases WPGP but weakens the utilitarian value→WPGP path; autonomy has no detectable role. Practically, emphasizing hedonic value is recommended when encouraging green behaviours involving premiums, particularly under high psychological costs; if WPGP cannot be raised, emphasizing utilitarian value may be preferable. Future research should use nationally representative samples, experimental/longitudinal designs to infer causality, and examine sociocultural contexts that condition psychological costs and value perceptions.

The sample is not nationally representative (younger, more educated, higher income, more female), limiting generalizability. The cross-sectional design precludes causal inference. Autonomy exhibited only acceptable (lower) reliability and showed no effects; further refinement of this construct may be warranted.