The Impact of Green Corporate Social Responsibility and Greenwashing on Green Packaged FMCG Product Purchase Intention: The Mediating Role of Perception

2.1 Grounded theories and main concept

2.1.1 The S–O–R

The Stimulus–Organism–Response (S–O–R) model was formally conceptualised and published in 1974 by Mehrabian and Russell, representing a significant extension of Woodworth’s (1929) earlier Stimulus–Response (S–R) model. Mehrabian and Russell posited that emotional and cognitive states of an individual (O – Organism) function as critical mediating variables, serving as the vital linkage between external environmental stimuli (S – Stimulus) and following behavioural responses (R – Response) exhibited by the consumer.

The framework comprises three core components:

(1) Stimulus: external sets off that initiate behavioural processes. Although stimuli represent necessary roots, they are not sufficient in themselves to extract behaviour; their ultimate effect is contingent upon internal organismic factors. Stimuli may include any object, event, or environmental cue capable of extracting attention or prompting action.

(2) Organism: the internal processing system of the individual, including cognitive evaluations, affective responses, interpretive processes, and other psychological states activated by the stimulus. This component represents the black box of human information processing and emotional experience that transforms external inputs into meaningful internal representations.

(3) Response: the observable behavioural outcomes or approach/avoidance tendencies manifested by the individual, which are profoundly shaped by preceding cognitive and affective organismic states.

The S–O–R framework has been widely adopted and adapted in prior consumer behaviour scholarship across varied contexts, proving instrumental in clarifying the mechanisms through which antecedent factors influence consumers’ purchase intentions and actual purchasing behaviour.

2.1.2 Conceptualisation of purchase intention towards FMCG packaged in green packaging

Behavioural intention is defined as the degree of effort an individual is prepared to exert in order to perform a specific action [11]. Building upon this foundation, green purchase intention refers to the likelihood that a consumer will select and purchase environmentally friendly products on the basis of their environmental knowledge and values, reflecting their willingness to pay a premium for goods and services offered by firms demonstrating sustainable practices [12]. Chen and Chang [13] further specified green purchase intention as the extent to which consumers are prepared to favour and pay more for products perceived as environmentally benign and congruent with their ecological concerns. Concurring with this perspective, Ghazali et al. [14] contended that green purchase intention captures an individual’s propensity or readiness to choose green, sustainable alternatives over conventional products manufactured using traditional materials and processes. Accordingly, purchase intention towards FMCG packaged in green packaging can be characterised as the degree of willingness exhibited by consumers to pay a premium for FMCG items encased in environmentally sustainable packaging materials, as opposed to comparable products utilising conventional, non-sustainable packaging.

2.2 Hypotheses development

2.2.1 GW and purchase intention towards FMCG packaged in green packaging (GPPI)

GW is defined as a misleading marketing strategy in which firms make false or exaggerated claims regarding the environmental sustainability of their production processes, products, or services [15]. Such practices generate consumer disbelief and prompt avoidance of ostensibly green offerings as a defensive response to perceived corporate duplicity [16]. Furthermore, Nguyen et al. [4] demonstrated that GW extracts negative affective reactions towards the firm, thereby diminishing consumers’ green purchase intentions. Accordingly, the following hypothesis is proposed:

H1: GW exerts a direct and negative effect on purchase intention towards FMCG packaged in green packaging.

2.2.2 GCSR and purchase intention towards FMCG packaged in green packaging (GPPI)

GCSR manifests through clear environmental stances and actions undertaken by firms [17]. GCSR activities are regarded as one of the most effective means by which organisations can satisfy burgeoning demand for sustainable consumption in an increasingly eco-conscious marketplace [18]. Empirical evidence provided by Ali and Sohail confirms the pronounced influence of CSR initiatives on green purchase intentions [19], as consumers perceive such actions not only as indicators of product quality but also as opportunities to express environmental concern through their purchasing behaviour [20]. Accordingly, the following hypothesis is advanced:

H2: GCSR exerts a direct and positive effect on purchase intention towards FMCG packaged in green packaging.

2.2.3 GW and PPR, EA, and PGP

GW heightens consumer disbelief, undermines motivation for sustainable consumption, and consequently reduces individuals’ inclination to engage in pro-environmental behaviour [21]. Consumers may perceive their own efforts as useless when confronted with insincere corporate practices, leading to diminished personal environmental responsibility [22, 23]. With respect to EA, individuals who detect GW tend to exhibit reduced belief in corporate environmental efforts, resulting in a weakened overall environmental concern [24, 25]. Szabo and Webster [26] further argued that GW diminishes perceived urgency of environmental issues by causing consumers to undervalue genuine ecological benefits.

Regarding PGP, GW adversely affects consumers’ ability to distinguish authentic sustainable packaging from misleading claims, particularly when knowledge is limited [27]. Some firms exploit green-coloured packaging to create illusory sustainability [28], whereas authentic green packaging mitigates GW perceptions and enhances consumer evaluation [29]. Based on the foregoing arguments, hypotheses are posited as follows:

H3: GW exerts a direct and negative effect on PPR.

H4: GW exerts a direct and negative effect on EA.

H5: GW exerts a direct and negative effect on PGP.

2.2.4 GCSR and PPR

In the context of a growing green consumption market, consumers are highly receptive to firms’ environmental initiatives, which encourage interaction with sustainable offerings [30]. Beyond implementing societal initiatives, GCSR actively stimulates consumers’ own sense of social and environmental responsibility [31]. Huang et al. [32] observed that individuals are more likely to enact personal environmental responsibility when supported by convincing CSR programmes. Accordingly, the following hypothesis is offered:

H6: GCSR exerts a direct and positive effect on PPR.

2.2.5 PPR and purchase intention towards FMCG packaged in green packaging (GPPI)

Individuals exhibiting strong environmental altruism demonstrate greater willingness to pay premiums for green products [33]. Nguyen Quoc et al. [34] similarly found that sustainable purchase intentions are significantly shaped by personal environmental responsibility, with highly responsible consumers displaying a marked preference for sustainable alternatives [9]. This behavioural tendency stems from the belief that individual green actions contribute meaningfully to environmental protection. Prior research has also established that PPR is negatively affected by GW [22, 23] yet positively influenced by authentic GCSR initiatives [32]. Consequently, the following hypotheses are advanced:

H7: PPR exerts a direct and positive effect on purchase intention towards FMCG packaged in green packaging.

H8a: PPR mediates the relationship between GW and purchase intention towards FMCG packaged in green packaging.

H8b: PPR mediates the relationship between GCSR and purchase intention towards FMCG packaged in green packaging.

2.2.6 EA and purchase intention towards FMCG packaged in green packaging (GPPI)

EA manifests in varied pro-environmental behaviours, including the preferential consumption of organic and sustainable products [35]. Rising demand for environmentally friendly offerings exemplifies consumers’ heightened ecological awareness [9]. Le et al. [36] reported that purchase decisions are increasingly influenced by environmental considerations, leading consumers to prioritise sustainable alternatives. EA has thus emerged as a hardy predictor of green consumer behaviour [37, 38]. On this basis, hypotheses are posited as follows:

H9: EA exerts a direct and positive effect on purchase intention towards FMCG packaged in green packaging.

H10: EA mediates the relationship between GW and purchase intention towards FMCG packaged in green packaging.

2.2.7 PGP and purchase intention towards FMCG packaged in green packaging (GPPI)

Product packaging serves both promotional and waste-generation functions, with conventional packaging constituting a substantial environmental burden [15]. Green packaging employs sustainable materials and energy-efficient production techniques [39]. Amid growing environmental concern, consumers associate sustainable packaging with superior quality and view its selection as an expression of ecological consciousness [40]. Empirical studies confirm that green or sustainable packaging significantly enhances green purchase intentions [41]. Drawing on these insights, hypotheses are formulated as follows:

H11: PGP exerts a direct and positive effect on purchase intention towards FMCG packaged in green packaging.

H12: PGP mediates the relationship between GW and purchase intention towards FMCG packaged in green packaging.

2.3 Research model

Based on the above arguments, the authors propose the following research model in Figure 1.

4.1 Results

4.1.1 Descriptive statistics

Following the data collection process, a total of 450 questionnaires were returned, of which 308 responses were deemed valid and retained for analysis (effective response rate: 68.4%). Respondents were categorised according to key demographic characteristics, including gender, age, educational attainment, and monthly income. The descriptive profile of the final sample (n = 308) is presented in Table 2.

Table 2. Demographic profile of respondents (n = 308)

Males comprised the majority of the sample (58.8%, n = 181), whilst females explained 41.2% (n = 127). In terms of age distribution, the two largest groups were respondents aged 55 and above and those aged 35 to under 45 (both 24.0%, n = 74), followed by the 45–under 55 group (18.5%, n = 57) and the 25–under 35 group (17.9%, n = 55). The youngest groups (18–under 25 and under 18) represented 8.4% and 7.1%, respectively.

Regarding educational attainment, the highest proportion held upper secondary qualifications (34.4%), followed by those with less than below upper secondary education (29.2%), bachelor’s degrees (21.4%), vocational/associate degrees (12.3%), and postgraduate qualifications (2.6%).

With respect to monthly income, the modal category was 7.5–under 15 million VND (34.4%, n = 106), followed by incomes surpassing 15 million VND (20.1%, n = 62) and the 4.5–under 7.5 million VND (18.8%, n = 58). Lower-income segments (below 1.5 million, 1.5–under 3 million, and 3–under 4.5 million VND) constituted 10.1%, 9.1%, and 7.5% of the sample, respectively.

Overall, the sample exhibits sufficient diversity across demographic dimensions, thereby enhancing the applicability of the findings within the Vietnamese FMCG consumer population.

4.1.2 Common method bias

The CMB test in Table 3, using Harman’s single factor method, showed that a single factor accounted for only 33.665% (< 50%) of the total extracted variance; therefore, the data did not exhibit general methodological bias [49].

Table 3. Common method bias

4.1.3 Cronbach’s alpha

The upshot of the reliability analysis is presented in Table 4. All constructs exposed Cronbach’s alpha coefficients surpassing 0.80, with corrected item-total correlations ranging from 0.570 to 0.778 (all > 0.30). These values confirm that the measurement scales demonstrate high internal consistency and that all observed variables are suitable for retention in the following analyses.

Table 4. Cronbach’s alpha reliability coefficients

4.1.4 EFA

EFA was performed separately for the independent, mediating, and dependent variables. Table 5 shows that eigenvalues exceeded 1.0 for each extracted factor.

Table 5. Total variance extracted of the factors

As shown in Table 6, the Kaiser-Meyer-Olkin (KMO) value is 0.898 (all > 0.50), and Bartlett’s test of sphericity was significant at p < 0.000 for all variable groups, confirming the appropriateness of factor analysis. All factor loadings were greater than 0.70, with no evidence of cross-loading or empty variables, thereby supporting unidimensionality and strong statistical significance of the measurement structure.

Table 6. Results of EFA

4.1.5 CFA

CFA was conducted to assess composite reliability (CR), average variance extracted (AVE), convergent validity, and discriminant validity. In Figure 2, the measurement model exhibited satisfactory fit: χ²/df = 1.790 (< 3.0), p < 0.001, CFI = 0.951 (> 0.90), TLI = 0.942 (> 0.90), and RMSEA = 0.051 (< 0.08). The GFI value of 0.899 fell marginally below the conventional threshold of 0.90 recommended by Hair et al. [50]. However, given that GFI is sensitive to sample size and number of observed variables, strict compliance to the 0.90 cut-off is often discouraged [51]. Following Baumgartner and Homburg [52], a GFI ≥ 0.80 is considered acceptable; thus, the model fit is deemed sufficient overall. All following structural analyses are therefore based on this validated measurement model.

Figure 2. Results of CFA

Table 7, the evaluation of composite reliability and validity yields the following results: composite reliability (CR) coefficients for all constructs exceed the recommended threshold of 0.70; AVE values are uniformly greater than 0.50; the maximum shared variance (MSV) for each construct is lower than its respective AVE. In Table 8, the square root of the AVE for every construct surpasses its correlations with all other constructs. These findings provide compelling evidence of strong convergent and discriminant validity. Accordingly, the measurement model is deemed psychometrically sound, and the analysis proceeds to the following stage of SEM to test the hypothesised relationships [30].

Table 7. Summary of Composite Reliability and convergent

Table 8. Discriminant validity results (Fornell and Larcker)

4.1.6 SEM

In Figure 3, the results of the SEM model analysis have a Chi-square value of 470.812 with 220 degrees of freedom (df = 220). Other statistical indicators are all satisfactory and statistically significant, Chi-square/df = 2.140 < 3; p = 0.000, showing that the observed variables are statistically significant at a good level. In addition, GFI = 0.879 > 0.8; CFI = 0.928 > 0.9; TLI = 0.917 > 0.9 and RMSEA = 0.061 < 0.08. Based on the analysis results, the author concludes that the model is compatible and can forecast the practical market.

The standardised path coefficients and hypothesis testing outcomes are summarised in Table 9. Of the nine direct-effect hypotheses, eight were supported at p < 0.05. GW (GW) exerted significant negative effects on PPR (β = -0.389, p < 0.001), EA (β = -0.570, p < 0.001), and PGP (β = -0.522, p < 0.001), thereby supporting H3, H4, and H5. Conversely, GCSR positively influenced PPR (β = 0.308, p < 0.001), supporting H6. EA (β = 0.233, p < 0.001), PPR (β = 0.176, p = 0.006), and PGP (β = 0.458, p < 0.001) all positively predicted GPPI, supporting H9, H7, and H11, respectively. GW also directly and negatively affected GPPI (β = -0.233, p = 0.008), supporting H1. However, the direct path from GCSR to GPPI was non-significant (β = -0.038, p = 0.509); thus, H2 was rejected.

Figure 3. SEM results

Table 9. Standardised regression coefficients and adjusted R²

Note: GW = Greenwashing; GCSR = Green Corporate Social Responsibility; EA = Environmental Awareness; PPR = Perceived Personal Responsibility; PGP = Green Packaging Perception; GPPI = Green Packaging Purchase Intention.

Table 10. Indirect effects and mediation testing results

Note: GW = Greenwashing; GCSR = Green Corporate Social Responsibility; EA = Environmental Awareness; PPR = Perceived Personal Responsibility; PGP = Green Packaging Perception; GPPI = Green Packaging Purchase Intention.

The adjusted R² values indicate substantial explanatory power of the structural model. Specifically, the antecedent constructs accounted for 32.5% of the variance in both PPR and EA (R² = 0.325 for each), 27.3% of the variance in PGP (R² = 0.273), and 67.1% of the variance in GPPI (R² = 0.671). These figures underscore the model’s hardy predictive capacity, particularly with respect to the ultimate dependent variable.

Mediation analysis using bootstrapping (5,000 resamples) further revealed significant indirect effects in Table 10. Although GCSR exerted no significant direct influence on GPPI, it positively affected GPPI indirectly through PPR (Standardised indirect effect = 0.054, p = 0.009), confirming PPR as a full mediator in the GCSR–GPPI relationship (only H8b supported). In contrast, GW negatively influenced GPPI indirectly via PPR (Standardised indirect effect = -0.068, p = 0.011; H8a supported), EA (Standardised indirect effect = -0.133, p = 0.006; H10 supported), and PGP (Standardised indirect effect = -0.239, p = 0.001; H12 supported). Thus, PPR, EA, and PGP parial mediated the respective relationships involving GW, providing empirical support for hypotheses H8a, H10, and H12. These findings highlight the pivotal role of consumers’ internal cognitive and affective processes in channelling the effects of corporate environmental signals—both genuine and deceptive—onto purchase intentions.

4.1.7 Multi-group SEM

According to Table 11, the Chidist p-value differs significantly between age and income groups (< 0.05). Therefore, the results indicate differences in the impact of the model's factors on GPPI [53].

Table 11. Multi-group SEM

Specifically, in Tables 12 and 13, the impact of GW→GPPI is quite strong in the following age groups: under 18, 18–under 25, and 25–under 35, and at most income levels. The impact of GCSR→GPPI is also quite strong in the following age groups: under 18, 18–under 25, and 25–under 35; and at the following income levels: below 1.5 million VND, 4.5–under 7.5 million VND, over 15 million VND. Furthermore, the impact of GW→PPR is quite strong in the following age groups: under 18, over 55, 35–under 45, 25–under 35; and in the following income levels: below 1.5 million VND, 1.5–3 million VND, 7.5–under 15 million VND, and 4.5–under 7.5 million VND. The impact of GW→EA is also quite strong in the following age groups: over 55, 45–under 55, 35–under 45, and 18–under 25, as well as in most income levels. The impact of GW→PGP is also quite strong in the following age groups: 18–under 25, over 55, 45–under 55, 25–under 35; and across income levels: over 15 million, 1.5–under 3 million, 4.5–under 7.5 million, and below 1.5 million. Additionally, the impact of GCSR→PPR is quite strong in the following age groups: 18–under 25, 25–under 35, and 35–under 45, and across income levels: over 15 million VND, 1.5–under 3 million VND, and 7.5–under 15 million VND. Meanwhile, the impact of PPR→GPPI is quite weak in most age and income groups, with the strongest impact in the 1.5–under 3 million VND. The impact of EA→GPPI is particularly strong in the following age groups: over 55, 25–under 35, and 45–under 55, as well as at income levels of 4.5–under 7.5 million VND, over 15 million VND, below 1.5 million VND, and 7.5–under 15 million VND. Finally, the impact of PGP on GPPI is quite strong in the age groups 45–under 55, 35–under 45, and over 55, as well as at income levels: over 15 million VND, 7.5–under 15 million VND, and the 4.5–7.5 million VND income group.

Table 12. Results of multi-group analysis by age

Table 13. Results of multi-group analysis by income

4.2 Discussion

The acceptance of H1 (β = -0.233; p < 0.05) indicates that contemporary consumers exhibit a heightened ability to detect corporate GW practices. Consequently, such misleading activities generate disbelief, negative emotions, and frustration, as consumers perceive their own pro-environmental efforts to be undermined. From a demographic perspective, this sensitivity to GW extends beyond Generation Z and younger groups. Within the income, all consumers displayed particular wariness, prioritising environmental considerations alongside personal needs during purchasing decisions. These findings are consistent of those reported by Meet et al. [15] and Sun and Shi [54].

Conversely, H2 was rejected (p-value = 0.509 > 0.05), consistent of Saif et al. [31] but unrelated from Huo et al. [8] and Le et al. [36]. Inconsistencies may stem from sampling differences—prior studies predominantly employed convenience sampling, potentially limiting representativeness—or from evolving consumer psychology. Demographically, the present sample comprised a substantial proportion of middle-aged and older respondents. Such consumers may prioritise tangible, practical product attributes over broader societal values, thereby attenuating the direct influence of perceived GCSR on purchase intention compared with younger consumers.

Support for H3, H4, and H5 underscores a growing consumer tendency to reject GW whilst elevating personal environmental responsibility, EA, and accurate perceptions of green packaging. Notably, the sample’s balanced age distribution, with the largest groups aged 35–under 45 and 55+, suggests that environmentally responsible consumption is no longer confined to younger demographics but has become more evenly distributed across generational groups. These results corroborate Sun and Shi [54] and Szabo and Webster [26].

Hypothesis H6 is accepted. This result also reflects the influence of GCSR on the intention to purchase FMCG packaged in green packaging. In other words, if businesses make positive contributions to society and the environment, it also strengthens and increases consumers' intention to purchase green products. From a demographic perspective, the higher the age and income, the more concerned consumers are about a company's social and environmental responsibility, in addition to concern about product quality. This research result is consistent with Saif et al. [31].

The acceptance of H7, H9, and H11 highlights that most of consumers with strong EA possess deeper ecological knowledge, heightened personal responsibility, and greater appreciation for authentic green packaging—all of which positively shape purchase intentions. Demographically, heightened awareness transcended age and educational boundaries, though higher-income respondents exhibited particular willingness to pay premiums for sustainable offerings, viewing such purchases as ethical investments in long-term environmental and personal health benefits. These observations are congruent with Carrión-Bósquez et al. [10], Meet et al. [15], and Shimul and Cheah [9].

Support for H8b confirms that PPR serves as a critical full mediator through which GCSR can influence purchase intention. The acceptance of H8b, H10 and H12 further demonstrates that GW distorts PPR, EA and PGPs, thereby indirectly suppressing green purchase intentions. Then, the above factors will play partial mediating roles in the relationship between GW—purchase intention. Notably, EA appears to be democratising across age groups—particularly evident among respondents aged 35–under 45, 45–under 55, and 55+; and educational levels, with even those holding below upper secondary qualifications exhibiting pronounced rejection of misleading practices.

In summary, this study contributes to the existing theoretical framework by expanding and supplementing the research models of previous studies [8-10] through the addition of cognitive factors (PPR, EA, and PGP) to better explain the mechanism of consumer green purchase intention formation. The study of intrinsic consumer factors as mediators also helps clarify and affirm the novelty of the research.

This work was supported by Industrial University of Ho Chi Minh City and University of Finance - Marketing. The authors have employed Grammarly software to check the spelling/grammar and polish the manuscript.