Consumers have become increasingly conscientious of their product choices. Often their preferences for certain products are grounded on moral considerations, as manifested in their willingness to pay more for products and services with strong ethical attributes (De Ferran and Grunert 2007). These “ethical consumers” support companies that adhere to value-based business practices (De Ferran and Grunert 2007), just as they avoid goods produced in questionable labour conditions (Andorfer and Liebe 2015). Consumers’ sensitivity to the environmental and societal implications of their consumption patterns often provides a base for their decision to consume ethically (Strong 1997).

One typical example of ethical consumption is the individual preference for fair trade (FT) products over non-FT commodities (De Pelsmacker et al. 2005). Nowadays, FT products are often sold alongside traditionally produced, non-ethical goods. With an increase in sales of FT products recorded in several major markets, such as the UK and the USA (Smithers 2014), Germany (Sarmadi 2015) and the Netherlands (GfK 2017), in the last five years, one can expect that the market for ethical goods will continue to thrive despite competition from more affordable non-ethical products.

Nonetheless, the fact that consumers have to pay more for FT products (Campbell et al. 2015; Castaldo et al. 2009) triggers the question on whether societal and environmental concerns would conflict with monetary considerations and self-interests. Recently, academic interest in the underlying mechanism behind FT product patronage has been surging, with researchers pursuing to identify relevant factors influencing people’s decision to purchase FT products.

The research described in this paper also aims at contributing to existing knowledge on FT product purchase motivations by testing the expanded version of the Theory of Planned Behaviour (TPB; Ajzen 1991) with the inclusion of moral obligation and self-identity to understand the factors shaping the decision to purchase FT products among conscious FT consumers in the Netherlands. The Netherlands has experienced a tripled increase in the number of households that buy FT products from 24% in 2007 to 69% in 2016 (GfK 2017). The decision to focus on conscious Dutch FT product consumers in this research is prompted by a finding that the determinants of FT product purchase intention differ among consumers and non-consumers – with attitude and subjective norm predicting the purchase intention of non-FT product consumers, while attitude and perceived behavioural control influencing FT consumers’ purchase intention (Ozcaglar-Toulouse et al. 2006). Moreover, as FT product consumers are less sensitive to price changes in FT products (Arnot et al. 2006), one wonders whether there are more important considerations that are taken into account when consumers decide to buy premium-priced FT products.

Although the extended TPB has been tested in previous studies into FT product purchase intention (e.g., Ozcaglar-Toulouse et al. 2006; Shaw and Clarke 1999), the extent to which the impact of the five TPB factors on FT product purchase intention differs between male and female consumers is not yet fully understood. This focus is necessary since knowledge of differences in factors that male and female consumers consider when buying FT products could aid in the design of an informed strategy of nudging those consumers to consistently patronize FT products by capitalizing on salient consumer beliefs and motivations.

Previous studies report that FT product purchase intention is higher among women than among men (Morrell and Jayawardhena 2010) and that women purchase FT products slightly more than men (Arnot et al. 2006), just as women are more willing to pay more for FT products than men do (Loureiro and Lotade 2005). The reported differences in FT product purchase behaviour between male and female consumers prompt the question on whether or not the factors influencing their FT product purchase decisions would also differ. As Bateman and Valentine (2010) claim, despite similarities in the behavioural intention and the actual behaviour of men and women, the decision-making processes preceding either their intention or behaviour might differ. And ethical decision-making processes between men and women, according to Morrell and Jayawardhena (2010), are anchored on different values.

Streaming from these claims, then, are the central research questions for this study:

  1. (1)

    To what extent do the extended TPB factors (attitude, perceived behavioural control, subjective norm, moral obligation, and self-identity) influence conscious Dutch FT consumers’ intention to purchase FT products?

  2. (2)

    To what extent do the effects of the extended TPB factors (attitude, perceived behavioural control, subjective norm, moral obligation, and self-identity) on FT product purchase intention differ among male and female Dutch consumers?

Theoretical Framework

Since its inception over 30 years ago, the TPB has been widely used to understand factors influencing the performance of various types of behaviour in different contexts. The theory evolved from the original Theory of Reasoned Action with the addition of perceived behavioural control as one of the predictors (aside from attitude and subjective norm) of behavioural intention (Fishbein and Ajzen 2010). TPB’s central thesis is that actual behaviour results from a person’s intention to perform that behaviour, which, in turn, is a function of that person’s attitude towards the behaviour, subjective norm, and perceived behavioural control (Ajzen 1991).

However, Fishbein and Ajzen (2010) also believe that TPB could still be expanded with the addition of personal norm or a sense of moral obligation and self-identity when studying behaviours with moral components and behaviours strongly tied to people’s roles or their social categories. De Ferran and Grunert (2007) claim that the inclusion of these two factors in the original TPB model increases its explanatory power. Indeed, as reported in various studies, the addition of both moral obligation and self-identity increased the variance explained for the intention to consume ethically (Dowd and Burke 2013; Shaw et al. 2000; Shaw and Shiu 2002).

The current study, hence, proposes that FT product consumers’ purchase intention is a function of their (i) positive attitude towards buying FT products, (ii) subjective norm, (iii) perceived behaviour control, (iv) moral obligation, and (v) self-identity. Moreover, this study looks into the extent to which those factors’ effects on purchase intention differ between male and female Dutch consumers.

FT Product Purchase Intention According to the Original TPB: the Effects of Attitude, Subjective Norm, and Perceived Behavioural Control

Attitude, defined as the “degree to which a person has a favorable or unfavorable evaluation or appraisal of the behavior in question” (Ajzen 1991, p. 188), triggers the performance of a behaviour in various contexts such as waste recycling (White and Hyde 2012) and charitable giving (Smith and McSweeney 2007). More importantly, the factor has been reported to increase people’s FT product purchase intention (Ozcaglar-Toulouse et al. 2006; Shaw et al. 2006).

Previous studies have also shown that perceived behavioural control, referring to the “ease or difficulty of performing the behavior” (Ajzen 1991, p. 188), matters in people’s decision to buy FT products. Such a perception of control, however, depends on consumers’ financial capability, as results of prior research have shown that FT product price impacts consumers’ decision to purchase those products (Shaw and Clarke 1999; Shaw et al. 2006).

Moreover, decisions on whether or not to engage in ethical actions are also predicated on social expectations. The impact of subjective norm or one’s perception of “social pressure to perform or not to perform the behavior” (Ajzen 1991, p. 188) on ethical consumption, specifically the purchase of FT products, has been found to be statistically significant in several studies (De Leeuw et al. 2014; Ozcaglar-Toulouse et al. 2006; Tarkiainen and Sundqvist 2005). Based on results of previous studies, the first three hypotheses are proposed.

  • H1: A positive attitude towards buying FT products increases Dutch consumers’ FT product purchase intention.

  • H2: Perceived behavioural control positively influences Dutch consumers’ FT product purchase intention.

  • H3: Subjective norms positively influence Dutch consumers’ FT product purchase intention.

Extending TPB with the Inclusion of Moral Obligations and Self-Identity

Moral norms or moral obligations are bound to influence the performance of behaviours with strong ethical dimensions (Conner and Armitage 1998). Schwartz (1977, in Schwartz and Howard 1980) introduced the concept of personal norm, which he conceptualized as an individual’s feeling of moral obligation to perform certain forms of prosocial behaviour, and underscored its impact on an individual’s proclivity to help another individual.

Ajzen (1991) regards moral norm (also known as perceived moral obligation; Rivis et al. 2009) as a “personal feeling of moral obligation or responsibility to perform, or refuse to perform, a certain behavior” (p. 199). The author admits that in certain contexts, it is not only social pressure that matters for the performance of a specific behaviour but also personal feelings of moral obligation to perform that behaviour.

Ethical consumption emerges from moral or ethical considerations (Bray et al. 2011) and an ethical consumer who strongly feels responsible towards society translates this feeling into the purchase of ethical products (De Pelsmacker et al. 2005). In the context of FT product purchase, one’s decision to buy an FT product is simply a decision between two alternatives, one of which is deemed more ethical than the other (Nicholls and Lee 2006). Previous studies have shown that people’s moral norms strongly predict their decision to buy FT products (Andorfer and Liebe 2015; Sunderer and Rössel 2012). These points prompt the fourth hypothesis.

  • H4: A feeling of moral obligation positively influences Dutch consumers’ FT product purchase intention.

Conner and Armitage (1998) argue that self-identity determines certain types of behavioural intentions. They define self-identity as “the salient part of an actor’s self which relates to a particular behavior” (p. 1444) and claim that it “reflects the extent to which an actor sees him- or herself as fulfilling the criteria for any societal role” (p. 1444).

From a social identity perspective, it has been noted that individuals who identify with certain roles or social categories will be inclined to perform behaviours that conform to those roles or categories (Fishbein and Ajzen 2010). Diverging views on the role of self-identity have been advanced. While several researchers underscore its direct impact of intention, as reported in a meta-analysis by Rise et al. (2010), it is also postulated that self-identity could moderate the impact of personal norm on intention (Fishbein and Ajzen 2010).

Since its inception over 30 years ago, TPB has been extended with the inclusion of self-identity as a predictor of behavioural intention (Sparks and Shepherd 1992), as people are expected to adjust their behavioural intentions when certain behaviours relate to issues that are central to their individual identities (Ozcaglar-Toulouse et al. 2006). Ethical behaviours, such as purchasing FT products, have identity components, as the social category of “ethical consumer,” for instance, is tied to a consumer’s idealized social image and it signifies the performance of an action that corresponds to such an image (Andorfer and Liebe 2013).

Self-identity or self-image concerns are surmised to impact individuals’ disposition to do what is considered right or moral to perform better than how they did in the past and to perform better than others, and to refrain from opportunistic acts (Teyssier et al. 2015). In a FT product purchase context, studies have shown that people’s self-identities (as ethical consumers) influence their propensity to continuously buy FT products (Andorfer and Liebe 2013; Ozcaglar-Toulouse et al. 2006). Emanating from these points is the fifth research hypothesis.

  • H5: The assumption of an identity (self-identity) as an ethical consumer positively influences Dutch consumers’ FT product purchase intention.

Differences in Ethical Consumption Between Men and Women

Reviewing the literature on gender and ethics, McCabe et al. (2006) noted that women tend to be more ethical than men, as the former are regarded more sensitive, emotional, and uncompetitive than the latter. Previous studies support this claim since women consider helping others more important than men do – an attitude that could lead to actual helping behaviour (Wilson and Musick 2006). Specifically, when compared to men, women are more inclined to do voluntary works (Taniguchi 2006; Themudo 2006), to spend more hours to volunteering (Manning 2010), and to be members of voluntary organizations (Themudo 2006). Moreover, the rate of charitable giving is higher among women than among men (Rooney et al. 2005), just as women tend to give more to charities than men (Mesch et al. 2011). These differences are often explained from a gender socialization standpoint (Bateman and Valentine 2010; Mason and Mudrack 1996), which emphasizes how men and women are socialized differently since “females are socialized to maintain relationships and empathize with others, while males are socialized according to a different set of principles” (Morrell and Jayawardhena 2010, p. 398).

More importantly, previous studies have reported that FT products purchase intention is higher among women than among men (Morrell and Jayawardhena 2010) and that women purchase FT products slightly more than men (Arnot et al. 2006), just as the former are more willing to pay more for FT products than the latter (Loureiro and Lotade 2005; Taylor and Boasson 2014). These findings lend credence to Hawkins’ (2012) contention of ethical consumption as a gendered act, as the role of consuming ethically largely falls on women, just as they are primarily targeted for ethical consumption and cause-related marketing initiatives.

Despite the supposed feminization of ethical consumption, as also evidenced by a research finding that individuals, regardless of gender, who consume ethically are viewed more feminine and less masculine (Shangi and Peloza 2016), male FT product consumers should not be regarded as an invisible segment, as the proportions of men and women who purchased different types of FT products are reportedly similar (Morrell and Jayawardhena 2010). Nonetheless, the reported differences in FT product purchase behaviour between male and female consumers also prompt the question on possible variations in the determinants of their purchase intention. As Bateman and Valentine (2010) contend, despite similarities in the intention and behaviour of men and women, the decision-making processes preceding either their intentions or behaviours might differ. More importantly, ethical decision-making processes between men and women, according to Morrell and Jayawardhena (2010), are anchored on different values.

For instance, the impact of attitude and perceived behavioural control on FT product purchase intention is moderated by gender, with attitude predicting male consumers’ FT product purchase intention and perceived behaviour control influencing female consumers’ purchase intention (De Leeuw et al. 2014). Although Leeuw et al.’s research did not find that gender moderates the effect of subjective norm on purchase intention, assuming that the effect of subjective norm on FT purchase would be stronger for women than for men may not really be baseless as women are reported to more easily yield to social influences (Eagly 1978) and to pay more attention to social cues than their male counterparts (Roberts 1991).

While nothing is still known about the extent to which consumers’ gender moderates the impact of moral obligation on ethical consumption, it has been reported in previous studies that women tend to be more ethical than men (McCabe et al. 2006) and that women are more likely to make ethical choices than men (Glover et al. 2002). Moreover, it is underscored that when consuming ethically, women might do so with a more intrinsic value than men (Starr 2009). Since an intrinsically motivated behaviour might be predicated on a personal norm or a feeling of moral obligation (Thøgersen 1996), it can be assumed that the effect of moral obligation on FT product purchase intention will be stronger for female than for male consumers.

Additionally, one experimental study shows that identity type interacts with consumers’ gender in influencing sustainable consumption (SC), as the impact of personal identity (defined as people’s view of themselves as individuals, hence conceptually similar to “self-identity”) on SC is significant for women but not for men, while the impact of social identity (defined as people’s view of themselves as part of a group) on SC is statistically significant for men but not for women (Costa Pinto et al. 2014). These theoretical insights prompt the hypothesis that:

  • H6: The effects of (a) attitude, (b) perceived behavioural control, (c) subjective norm, (d) moral obligation, and (e) self-identity on Dutch consumers’ FT product purchase intention are moderated by their gender.

Figure 1 shows the expanded version of the Theory of Planned Behaviour with the inclusion of moral obligation and self-identity, which will be tested to determine the factors influencing conscious FT product consumers’ intention to purchase FT products. The figure also shows the hypothesized moderating role of consumers’ gender in the relationship between the extended TPB factors and FT product purchase intention.

Fig. 1
figure 1

Conceptual model for the impact of extended TPB factors on FT product purchase intention and for the moderating role of consumers’ gender

Full size image

Method

Design and Procedure

Data used to test the research hypotheses were collected by a Dutch research agency for one week. The agency, with a panel that represents the Dutch population, sent a link to the online questionnaire to 785 panel members. Given the study’s focus on the mechanism behind FT consumers’ decision to purchase FT products, it was decided to exclusively focus on conscious FT product consumers, as they are known to be less sensitive to price changes in FT products and to base their FT product purchase decisions on considerations other than price (Arnot et al. 2006). Hence, a screening question (“How would you describe yourself?”) – with five options pertaining to the possible FT consumption behaviour of survey respondents – was asked right at the start of the survey.

Four hundred eleven (53%) respondents indicated that they consciously buy certain types of fair trade products, while 88 (11.21%) reported to be conscious consumers of all fair trade products. Hence, those 499 respondents were requested to complete the entire questionnaire and data they provided were subjected to statistical analyses.

On the contrary, 23 respondents (3%) indicated to have stopped buying FT products, 232 respondents (30%) purchased FT products without consciously thinking of their motivations, and 17 respondents (2%) have never bought FT products prior to the survey. Therefore, a total of 272 respondents were instructed to not complete the online questionnaire after answering the screening question.

Respondents

Of the 499 respondents whose data were used for statistical analyses, 232 were males (46.5%). Respondents’ age ranged from 18 to 78 years, with a mean age of 43.27 (SD = 14.55). Table 1 presents the complete demographic information of the survey respondents.

Table 1 Complete demographic information of the survey respondents
Full size table

In terms of their FT product purchase behaviour, most respondents normally bought FT products from general supermarkets (N = 392, 78.6%). In terms of FT product types that respondents often purchased, bananas (N = 144, 28.9%), chocolate and cocoa products (N = 119, 23.8%), and coffee (N = 107, 21.4) topped the list.

Measurements

The dependent variable “FT product purchase intention” was measured with two newly formulated statements. The first predictor “attitude towards purchasing FT products,” operationalized in this study as an individual’s perception of the positive impact of purchasing FT products, was also measured with two newly formulated items.

“Subjective norm” was measured with four items that were inspired by the scales of Shaw et al. (2000) and Ozcaglar-Toulouse et al. (2006), whereas “perceived behavioural control,” operationalized as the extent to which people believe that they have the financial means and the possibility to purchase FT products, was measured with two newly formulated items.

Two items, which were modified versions of statements by Sunderer and Rössel (2012), were used to measure “moral obligation.” Finally, self-identity was measured with three items, two of which were substantially reformulated versions of items by Sparks and Shepherd (1992). All the items used to measure the research constructs are presented in Table 3.

Table 2 reports the mean scores and standard deviations of the constructs for both male and female consumers. They were calculated to give an indication of their self-reported perceptions and beliefs. An independent sample t test was performed to see whether the scores for the various constructs significantly differ between male and female consumers. The results show that values for subjective norm, perceived behavioural control, and purchase intention are significantly higher for men than for women. That FT product purchase intention is higher for male consumers than for their female counterparts counters results of a study by Morrell and Jayawardhena (2010), as they found that female consumers expressed a higher intention to buy FT products than male consumers did.

Table 2 Mean and standard deviation values for the research constructs and results of the independent sample t test
Full size table

Results

Measurement Reliability and Validity

Prior to testing the structural model with data for both male and female respondents, requirements of instrument validity and reliability must be met. Confirmatory factor analysis using AMOS 22.0 was performed to determine the discriminant and the convergent validity of the scales. Hair et al. (2006) recommend a factor loading (FL) value higher than 0.50 for an item to be significant. Table 3 presents the factor loading values for the individual items. Additionally, at the construct level, Hair et al. (2006) proposed the calculation of composite reliability (CR) and average variance extracted (AVE) instead of Cronbach’s alpha when using structural equation modelling (SEM).

Table 3 Factor loading values for the different items measuring the research constructs
Full size table

CR values are adequate for all factors (> 0.60; Bagozzi and Yi 1988), just as AVE values are higher than 0.50 (Fornell and Larcker 1981); hence, the constructs have convergent validity. Inter-correlations among the seven constructs were also determined using correlation analysis before the structural model was tested. Values in Table 4 indicate that strong correlations (correlation values between .70 and .90; Burns and Burns 2008) among the constructs do not exist. To test for discriminant validity, all square roots of AVE need to be higher than the inter-construct correlation estimates. This criterion is fulfilled. Details for AVE as well as square roots of AVE, CR, and correlation values are presented in Table 4.

Table 4 Correlations among the research constructs for the overall sample (N = 499) with the square roots of the AVE values in the diagonal
Full size table

Fit of the Measurement Model

The measurement model was tested to determine its fit to the research data. Test of the fit of the measurement models for both original and extended TPB (for male and female consumers) indicates that the measurement models have an acceptable fit (see Table 5). Hence, our measurement models possess configural invariance.

Table 5 Model fit indices for the original and the modified versions of TPB
Full size table

For a meaningful comparison of the model of FT product purchase intention for both male and female consumers, the instrument measuring the various constructs must possess cross-gender equivalence. To meet the requirement of equivalence, configural and, at least partial, metric or scalar invariances must be confirmed to compare the findings for the two groups of consumers (Hair et al. 2006; Steenkamp and Baumgartner 1998; Vandenberg and Lance 2000).

Metric invariance was tested by means of nested multiple-group CFA. For the original TPB model, there is no significant difference in χ2 (7.33, df = 6, p = .29) between the free and the restricted model (i.e., factor loadings restricted to being equal across genders). Thus, metric invariance can be assumed. However, there is a significant difference in χ2 (19.30, df = 9, p = .02) between the free and the restricted model of the extended TPB model. However, a partial metric invariance model, in which one additional factor loading of the construct subjective norm was restricted to being equal, produces a non-significant change in χ2 (12.48, df = 8, p = .13) compared to the model in which the factor loading fixed at unity to define the scale is set equal. Hence, assuming partial metric invariance for the extended TPB model seems justified (Thøgersen et al. 2015).

Test of the Original TPB Model

Prior to the test of the extended TPB model, the original model with three predictors, namely, attitude, subjective norm, and perceived behavioural control, was first tested. Multi-group analysis of TPB model with data from male and female consumers shows an excellent model fit, as presented in Table 5.

By looking at the equality of structural weights, the significance of the overall difference in the factors influencing FT product purchase intention of both male and female consumers was determined. The path coefficients, as well as the results of the chi-square test of the change in fit as a consequence of constraining each structural weight to be equal across genders, are reported in Table 6.

Table 6 Standardized and unstandardized (in parenthesis) parameter estimates of the structural model (N = 499)
Full size table

This initial analysis reveals that male consumers’ FT product purchase intention is influenced by subjective norm (β = .56, p < .001) and attitude (β = .28, p < .001), while among female consumers, their purchase intention is predicated on the three TPB factors: perceived behavioural control (β = .36, p < .001), attitude (β = .34, p < .001), and subjective norm (β = .18, p < .05). The χ2 difference test shows that the path from subjective norm to purchase intention is significantly higher for male consumers than for female consumers. R2 values for the original model for both male and female consumers are pegged at 0.60 and 0.44, respectively.

Test of the Extended TPB Model

The extended version of the TPB model was then tested with the addition of moral obligation and self-identity. The modified model also has a good fit (see Table 5).

Analysis of the modified model indicates that male consumers’ FT product purchase intention is influenced by moral obligation (β = .31, p < .01), subjective norm (β = .27, p < .01), and self-identity (β = .25, p < .05). These results support hypotheses 3, 4, and 5. Among female consumers, their FT product purchase intention is also predicated on three factors: self-identity (β = .40, p < .001), moral obligation (β = .35, p < .001), and perceived behavioural control (β = .23, p < .001). Therefore, hypotheses 2, 4, and 5 are supported.

When looking at both the original and the extended TPB models (Table 6), only the χ2 difference test for the path between subjective norm and purchase intention is significantly different for male and female FT consumers in the Netherlands. This indicates that only the effect of subjective norm on FT product purchase intention is moderated by consumers’ gender. Therefore, only hypothesis 6c is supported.

The addition of moral obligation and self-identity to the original TPB model substantially increased the R2 values for the models tested with data from male consumers (R2 = 0.70) and from female consumers (R2 = 0.70). These findings support what is already known in previous studies that expanding TPB with moral obligation and self-identity can increase the model’s explanatory power (De Ferran and Grunert 2007; Dowd and Burke 2013; Shaw et al. 2000; Shaw and Shiu 2002).

Discussion of Results, Implications, and Future Research Directions

Discussion of Results

Test of the original TPB with data from 499 FT product consumers in the Netherlands reveals that attitude, subjective norm, and perceived behavioural control significantly influence those consumers’ FT product purchase intention (except for male consumers in which perceived behavioural control has no effect at all). However, the inclusion of moral obligation and self-identity rendered the impact of attitude and subjective norm (specifically, for female consumers) insignificant. The insignificant effect of subjective norm on ethical consumption when moral obligation and self-identity were added to the model corroborates results of previous studies (Shaw and Shiu 2002; Shaw et al. 2000), which reported that the inclusion of the two variables reduced the effect of attitude. In another research, even, the inclusion of moral obligation alone to TPB when used to study pro-environmental behaviour reduces the effects of attitude and subjective norm (Harland et al. 1999).

In the current study, attitude’s impact on FT product purchase intention was obliterated with the inclusion of two the constructs. In situations involving decisions to act ethically, one can only speculate that moral obligation might already suffice to prompt the performance of such an action, which eventually reduces the role of a more general attitude construct as a determinant of behavioural intention. As respondents for this study are conscious FT product consumers, the recurrence of the purchase behaviour might have resulted in the internalization of the act as a personal norm or moral obligation (Olander and Thøgersen 1995). When viewed in the context of ethical behaviour performance, the removal of attitude’s significant effect when moral obligation is added into the model might also suggest that one’s feeling of moral obligation can replace attitude when understanding people’s inclination to engage in an action with a strong moral dimension (Chan and Bishop 2013).

It is worth noting that the introduction of moral obligation and self-identity did not obliterate the impact of subjective norm on male consumers’ purchase intention, although that construct’s effect on intention vanished for female consumers. This is an interesting outcome as women are described to be more receptive to social influence than men (Fischer and Arnold 1994). However, in the context of prosocial behaviour, it has been noted that accounting for personal norms would render the direct effect of subjective social norms on the behaviour insignificant (Thøgersen 2006) and that the impact of subjective norm on behavioural intention might even be mediated by personal or moral norms (Bamberg and Möser 2007). As moral obligation or personal norms “are assumed to be (superficially) internalized social norms” (Thøgersen 2006, p. 256), it can be surmised that norms regarding FT product purchase are more internalized among female than male consumers.

An implication of this finding, hence, is that in situations when decisions regarding the performance of an ethical action are hinged on one’s capacity to engage in the act, women are more likely to consider the resources they have when deciding whether or not to purchase FT products instead of succumbing to the expectations of their social contacts.

Results of this study clearly indicate that the impact of perceived behavioural control on FT product purchase intention is significant for female consumers only but not for their male counterparts. This finding supports what is known in another study that perceived behavioural control matters more for female consumers’ FT product purchase decisions than for male consumers (De Leeuw et al. 2014). As FT products cost more than non-FT products (Pedregal and Ozcaglar-Toulouse 2011), the impact of financial considerations on FT product purchase decisions would probably be salient. Thus, even if women are more likely to behave ethically than men, women’s proclivity to act ethically, for instance, by buying ethical products, might also depend on whether or not they have the means to do so.

Previous research showed that women tend to buy FT products more than men (Sunderer and Rössel 2012). Nevertheless, results of the current study indicate that the purchase intention for men is higher than for women. Though this difference might be attributable to different sample characteristics given the research’s focus on FT product consumers, it can also be assumed that since women tend to do the groceries more than men (Dholakia 1999), the former might be more aware of prices differences between FT and non-FT products than the latter.

This awareness of price differences might also explain why perceived behavioural control matters more for female than for male consumers when deciding whether or not to purchase FT products, as female consumers’ decision to buy FT products might also be weighed against the need to buy other non-FT but necessary daily goods within the bounds of their allocated grocery budget. It should be noted that in this research, female consumers’ score for perceived behavioural control (PBC) is significantly lower than that of their male counterparts.

For male consumers, however, perceived behavioural control has no effect on FT product purchase intention, which suggests that monetary considerations have no bearing on their intention to buy FT products. This is somehow supported by their mean scores for PBC, which indicate that male consumers are in some ways confident of their ability to buy FT products.

Instead, an important determinant of male consumers’ FT product purchase intention is their belief in what individuals within their social networks expect them to do. However, as the conceptualization of social influence in this study is only limited to people’s belief in how their significant others expect them to behave, one cannot fully claim that the result challenges the mainstream view of women’s susceptibility to social influence.

Nonetheless, the findings that subjective norms matter only for male consumers but not for their female counterparts and that perceived behavioural control matters only for female consumers but not for their male counterparts seem to suggest that the impact of subjective norm on behavioural intention depends on people’s estimation of their capacity to engage in an action. The impact of subjective norm on male consumers’ FT product purchase intention might be an indication of an attempt to conform to their social contacts’ expectations based on the notion that they ought to consume ethically because they have the financial power to do so.

As expected, a feeling of moral obligation contributes to both male and female consumers’ FT product purchase intention, and this finding echoes results of previous studies into the impact of the factor on ethical consumption. This is not surprising since the behavioural intention of interest has a strong moral dimension, which is bound to trigger consumers’ estimation of the ethical desirability of performing the behaviour.

The influence of self-identity on FT product purchase intention is also statistically significant for both male and female consumers. People who view themselves in a certain way will most likely behave in a manner that corresponds to their desired self-image. Ashforth et al. (2008) convincingly established the link between an individual’s identity and behaviour by asserting that people’s awareness of their identities results in their acceptance of values, goals, and beliefs associated with those identities, which eventually induces them to engage in certain actions that mirror their identities. Since research respondents are already FT product consumers, it is also likely that they view their decision to continue patronizing FT products as an opportunity for them to fully assume their identities as socially and environmentally conscious individuals (Connolly and Shaw 2006).

Implications for Policy and Practice

An important implication of these findings for practice is that FT product purchase intention is mainly driven by identification processes. FT is an attribute that distinguishes one brand from its competitors, creating an added value to the brand. Consumers perceive the brand as a valuable object through which they can connect to their identities and even enhance their self-concepts. This can lead to a close bond between the brand and consumers, which a company needs to maintain. A relationship between consumers and company based on identification processes has an enormous potential for generating positive long-term affective and behavioural outcomes (Curras-Perez et al. 2009).

Given the strong impact of a feeling of moral obligation on male and female consumers’ intention to purchase FT products, and the fact that such feeling of obligation emerges from a moral belief (Nelson et al. 2006), the pivotal role of information should be tapped. This assertion is based on Fishbein and Ajzen’s (2010) thesis that beliefs are anchored on what people know about the behaviour and provision of new information about a behaviour could alter beliefs and might effectively modify intentions and actions. Additionally, it is reasonable to emphasize in the presentation of FT goods and in FT product information dissemination the desirable consequences of an individual’s decision to patronize FT products, as the development of personal norms is also predicated on people’s awareness of the consequences of their actions for others (Thøgersen 2004).

As one study has shown (De Pelsmacker and Janssens 2006), information plays a critical role in shaping consumers’ attitude towards FT products and their intention to purchase them. Maintaining the feeling that buying FT products is the right thing to do needs to be pursued by increasing certain values (Ortberg et al. 2001), especially those pertinent to FT product purchase (e.g., “buying FT products ensures fair treatment of farmers in developing countries”). Once positive beliefs about FT product purchase have been identified, the most appropriate interventions could be designed (Steinmetz et al. 2016) to effectuate change in the public’s consumption patterns and to prompt them to make ethical consumption choices.

Efforts to increasingly inform, educate, and persuade a large segment of consumers could take a multi-sectoral approach by involving relevant institutions (e.g., government agencies, consumer groups, non-profit organizations) in a widespread campaign to highlight the societal and the environment benefits of purchasing FT products. Behavioural-change interventions such as information provision, persuasion, and motivation are reported to have significant effects on people’s attitude towards a desired behaviour, on their perception of subjective norms, and on their assessment of whether or not they are able to perform a behaviour (Steinmetz et al. 2016). Hence, strategic use of the interventions previously mentioned could be explored to strengthen the impact of the extended TPB factors on consumers’ willingness to purchase or to continue purchasing FT products.

Future Research Directions

Despite the study’s potential to offer relevant insights into Dutch FT product consumers’ intention to purchase FT products, certain aspects of the research (e.g., research model, measurement) could still be improved in future research. First, although previous studies measured attitude towards a behaviour on a semantic differential scale with contrasting pairs of concepts such as “good-bad” and “harmful-beneficial,” as originally proposed by Ajzen and Fishbein (1980), the attitude construct in the present study was measured with items that emphasize the positive attributes of purchasing FT products, which capture specific instead of overall evaluations of the behaviour of interest. Specifically, attitude, as operationalized in this study, pertains to the perceived beneficial impact of the action considered good instead of simply assessing it as good, positive, or desirable (the more general evaluation of the behaviour).

The way attitude was measured in this study is somewhat similar to De Pelsmacker and Janssens (2006) treatment of attitude as either support for FT or skepticism towards FT. This approach of measuring attitude, however, might have critical implications for the findings, as viewing FT product purchasing a good thing is not similar to believing in the possible positive implications of buying FT goods. Hence, despite findings that TPB, when used to study behaviours with strong moral components, can be modified by replacing the attitude construct with moral obligation (Chan and Bishop 2013), the distinct effects of both constructs could still be investigated by ensuring that attitude is measured as conceptually different from moral obligation, specifically with items that do not overlap with those supposedly measuring moral obligation (Manstead 2000). The second point is critical since attitude towards an act with a moral consequence for another, when measured with the instrument Ajzen and Fishbein (1980) proposed, measures a personal norm or a feeling of moral obligation (Thøgersen 1996).

Second, the attitude construct in the study is limited in its focus on evaluating an ethical consumption behaviour. As consumers are likely to purchase ethically when they do not have to pay more for a product or to compromise product quality over ethical considerations (Bray et al. 2011), research into FT product purchase should consider the impact not only of price but also of product quality. An important point for future research, therefore, is the emphasis on both FT product purchase behaviour and FT product as relevant targets for attitude.

Third, with the researchers’ decision to stick to the original operationalization of subjective norms (expectations of what people regard as appropriate), the effect of the other dimension of social influence was not taken into account. Cialdini and Goldstein (2004) differentiated social influence into two types, namely injunctive norms (perceptions of what others approve) and descriptive norms (perceptions of what others do). As it is known that people tend to consume ethically when people around them do the same (Starr 2009), the model tested in this study could be further extended by replacing “subjective norms” with a two-dimensional social influence construct and to test whether or not the effects of the two dimensions on FT product purchase intention would substantially differ between male and female consumers.

Fourth, the present study also found that perceived behavioural control is a critical consideration for female consumers’ decision to purchase FT products. Currently, the operationalization of PBC mostly pertains to consumers’ assessment of their financial capacity to buy FT products. Predicated on the view of perceived behaviour control as “a function of beliefs about resources, opportunities, and other factors that facilitate or obstruct behavioral performance” (Yzer 2012, p. 104), measuring the construct should emphasize not only an individual’s financial capacity to purchase premium-priced FT products but also the convenience of making the purchase and the locational accessibility of those products for consumers. The emphasis on convenience could still be specified by highlighting the impact of the location of shops that sell FT products, as location has been noted a critical issue in the decision to purchase FT products, even among FT consumers (Becchetti and Rosati 2007).

Although previous studies measure PBC by emphasizing the possibility (impossibility) for consumers to buy FT products (De Leeuw et al. 2014) and of the difficulty in purchasing those products (Shaw et al. 2000), the conceptualization only partly captures the concept of behavioural control. In a FT product purchase context, therefore, perceived behavioural control needs to be adequately conceptualized as a two-dimensional construct (price and access to as well as availability of products) and future research could consider examining whether the two PBC dimensions would have different effects on FT product purchase intention.

Fifth, as feelings of moral obligation or personal norms play a pivotal role in increasing both male and female consumers’ intention to purchase FT products, understanding how personal norms develop should be a worthwhile research pursuit. From a theoretical standpoint, focusing on the development of personal norms in the context of FT product purchase would provide an in-depth insight into how those personal norms evolved. Such an insight is consequently valuable from a practical standpoint since knowing “how personal norms are formed, changed, and are activated is potentially useful for the design of social interventions to promote prosocial behaviors” (Thøgersen 2002, p. 882), such as purchasing ethically produced goods.

Bandura’s (1991) Social Cognitive Theory of Moral Thought and Action stipulates that people’s moral standards partly emerge from their social environment. Additionally, Cialdini’s (2007) “social proof” concept underscores that an individual’s appraisal of the correctness of something is based on other people’s estimation of that thing’s correctness. Moreover, a study by Van der Werff et al. (2013) found that people who assume strong environmental identities have intense feelings of moral obligation to act in an environmentally friendly manner. Future research, therefore, could focus on the potential contributions of both social norms and individual identities in enhancing consumers’ feelings of moral obligation to purchase FT products.