Making the Customer-Brand Relationship Sustainable: The Di ﬀ erent E ﬀ ects of Psychological Contract Breach Types on Customer Citizenship Behaviours

: Customer citizenship behaviours are important for a sustainable customer-brand relationship, yet little is known about the e ﬀ ect of psychological contract breach on citizenship behaviours. From the perspective of psychological contract theory, this research examined the impact of psychological contract breaches on customer citizenship behaviours through the mediating mechanism of psychological contract violation in the customer-brand relationship. Experiments were used to assess the e ﬀ ect of two types of psychological contract breaches on customer citizenship behaviours. The results show that the negative e ﬀ ect of a relational contract breach di ﬀ ers between the dimensions of citizenship behaviours. A relational contract breach has direct and indirect negative e ﬀ ects on recommendation behaviours and helping behaviours, but it does not a ﬀ ect voice behaviours. A transactional contract breach directly and negatively inﬂuences recommendation behaviours, but it has no e ﬀ ect on voice behaviours. A transactional contract breach indirectly increases helping behaviours through psychological contract violation. Psychological contract violation partially mediates the e ﬀ ect of a transactional contract breach on customer recommendation behaviours and fully mediates the e ﬀ ect of a relational contract breach on helping behaviours. These ﬁndings generate managerial implications for ﬁrms aiming to maintain sustainable customer-brand relationships. ect of a relational contract breach was di ﬀ erent between dimensions of customer citizenship behaviours: it had a direct signiﬁcant negative e ﬀ ect on recommendation behaviours, but no direct e ﬀ ect on voice and helping behaviours. We did observe a direct positive e ﬀ ect of relational contract breaches on voice behaviours. It is possible that customers want to express their negative experiences and provide suggestions to the brand. The e ﬀ ect of the relational contract breach on recommendation and helping behaviours was mediated by participants’ perception of psychological contract violation. The results indicate that relational contract breach led to a negative emotional experience, and these emotional experiences decreased recommendation behaviours


Introduction
Building a sustainable relationship with consumers is the biggest challenge for brands. Past research on the bilateral brand-consumer relationship has mostly focused on the passive role of customers; however, recent literature has proposed that customers are not passive respondents but active participants [1]. With the development of information technology and the popularity of smartphones, customers are able to share their experiences of using products with others and propose suggestions to brands, interact with similar customers in online brand communities and so on. Such behaviours are called consumer citizenship behaviours [2]. Citizenship behaviour is not a requirement for a successful product or service delivery, but it has positive effects on brand image and brand relationships [3]. Recent research on the interactive process of the consumer-brand relationship has typically explored it from two perspectives to acquire an in-depth understanding of this phenomenon [3][4][5].
Psychological contract theory offers a unique view to explain the bilateral reciprocal customer brand relationship [6]. Consumers form beliefs about the reciprocal obligations between themselves and the brand, and these obligations are based on perceived promises and expectations [7]. Psychological contracts represent consumers' structured knowledge about the patterns of resource exchange between two parties in a relationship. During the relationship-building process, consumers first form expectations from the brand's advertisement or consumption norms. Consumers and brands interact with each other through the steps of consumption. Consumers may feel satisfied with and trust a brand when it fulfils their expectations. Some consumers even feel grateful to the brands [3] and positively recommend them to other consumers. However, when brands violate expectations, these high-commitment consumers (compared with low-commitment consumers) show stronger negative changes in attitude [8]. Healthy relationships between consumers and brands are often disrupted by events such as brand transgressions, product harm crises, and service failures. For example, Tesla promised a Chinese customer the opportunity to be the first Tesla car owner, but the promise was ultimately broken. The customer was deeply angry and destroyed the new Tesla car in an act of revenge.
When consumers perceive that a brand has not fulfilled a commitment, they may feel betrayed and behave negatively toward the company. Their relationship may reach its endpoint: brand aversion [9]. Recent research has shown that consumer perceptions of violations in (vs. out of) contracts result in a negative attitude [8]. Whether a contract breach affects consumer citizenship behaviour and its process needs further theoretical exploration. Most of the published literature focuses on the antecedents of the positive side of consumer citizenship behaviours. For example, consumers have been found to perform citizenship behaviours when a company's reputation is good [10], the employees are benevolent [11], other consumers also perform citizenship behaviours [12], they have agreeableness and extraversion personalities [13] and so on. Only a scattering of research has explored the negative side, and the results have shown that service scripts have a negative effect on consumer citizenship behaviours when the service is performed by employees with low customer orientation [14]. Furthermore, there are different types of psychological contracts (relational contract and transactional contract): the relational dimension reflects the long-term, interactive component and the transactional contract involves the consumer's short-term economic expectation from the brand. The two types of contract breaches are known to have different effects on employees' organisational citizenship behaviour [15]. However, less is known about the different effects of the two psychological contract breaches on consumer citizenship behaviours.
With the aim of closing the above-mentioned gaps in the literature, this study addresses the above issues by exploring the impact of the two types of contract breaches on consumer citizenship behaviours. We demonstrate that different types of breaches (relational contract and transactional contract) have different impacts on customer citizenship behaviours. Thus, this research contributes to the literature on the customer-brand relationship and consumer citizenship behaviours. Most of the previous research on the psychological contract has focused on transactional contracts in the employee-company context. Moreover, we uncover an important antecedent that negatively influences consumer citizenship behaviours. Customers' active participation plays a more important role than ever before, and both their positive and negative participations have big effects on the quality of the customer-brand relationship. From a practical perspective, identifying negative antecedents of customer citizenship behaviours is very urgent for marketing managers who strive to maintain healthy customer relationships.
The remainder of this article is organised as follows. We first present our theoretical background, conceptualisation and hypothesis development. Next, the research methodology is presented, which includes the design of the study, research materials, and research results. The final section summarises the findings and offers managerial implications.

Psychological Contract and Psychological Contract Breaches
Consumers develop psychological contracts with brands and organisations. Psychological contracts in the employee-organisation relationship have been studied [16], and the relationship between consumers and brands is similar to the relationship between employees and companies [8]. For example, both types of relationships (brand-customer vs. employee-organisation) are formed between an individual and an organisation, and in both cases, implicit commitments or expectations exist between the individuals and the organisation. Consumers' psychological contract expectations are generally based on the commitments made by the brands and other consumers through mechanisms such as advertisements, sales staff product recommendations and word of mouth.
The classical psychological contract literature describes two types of consumer psychological contract: the transactional psychological contract and the relational psychological contract [17]. A recent paper suggested that there are four types of contracts: relational, standard, transitional and captive contracts [5]; however, in their definitions, the features of the standard contract are similar to those of the transactional contract. Therefore, we address the two classical contract types in this paper. The transactional psychological contract is built on the basis of short-term interest and rewards and is mainly concentrated on a specific and interactive economic relationship [17,18]. In contrast, the relational psychological contract involves long-term and affective interactive relationships that typically include high affective involvement, such as support for individuals and concern for families. Consequently, relational psychological contracts inspire a greater sense of participation and belonging than transactional ones [7]. The literature on contract types reports that the saliency of the employers' contract type affects employee behaviours. The importance that employees attach to transactional obligations has a negative effect on their commitment to their employer and the degree to which they fulfil those obligations; in contrast, the importance attached to relational obligations has a positive effect on employee behaviours [19].
A psychological contract breach occurs when customers perceive that the company has failed to fulfil its promises [20]. According to the definition, a consumer psychological contract breach is different from a product harm crisis. A product harm crisis is an incident in which the product has a shortcoming [21]. A product harm crisis occurs before the consumer perceives a psychological contract breach, which means that a crisis may result in the consumer's perception of a psychological contract breach. In other words, a product harm crisis is one of the antecedents of a psychological contract breach. A psychological contract breach refers to consumers' subjective feelings, while a product harm crisis is an objective description of an event. For instance, when dining at a restaurant, customer A and customer B have the same experience of waiting for more than 15 min before being served. Customer A feels disappointed, while customer B does not feel disappointed, possibly because customer B did not have an expectation of quick service. In other words, although both customers are faced with the same event, a psychological contract breach arises in customer A but not in customer B. Customer A feels that the restaurant failed to fulfil a promise (individual differences such as customer personality may affect customer satisfactions [13], and there may be other explanations for this phenomenon in this context). Whether an individual believes that contract breach has occurred is contingent on incongruent salience [17], which is the gap between the consumer's expectation of service quality and the consumer's perception of the delivered quality. A psychological contract breach decreases an individual's trust in and satisfaction with the organisation [22].

Psychological Contract Violation
Psychological contract violation refers to the feeling of anger and betrayal when an individual believes that an organisation has failed to fulfil its obligations [17]. In some psychological contract studies, the definition and measurement of 'psychological contract violation' have been similar to those of a psychological contract breach [23]. However, Morrison and Robinson (1997) asserted that the two concepts are different; psychological contract violation is an emotional and affective element constituted by disappointment, frustration, anger and distress [17]. Whereas a psychological contract breach is the cognitive result of an organisation's failure to fulfil its commitments. A psychological contract breach may lead to contract violation [20]. From the cognitive aspect, a psychological contract breach is followed by the emotional experience of contract violation. Psychological contract breaching and violation are different but related concepts. If an individual perceives a serious contract breach and attributes the breach to reneging, he or she will experience a strong feeling of contract violation [20].
Psychological contract violation can lead to various negative consequences. When employees perceive a contract breach and have feelings of violation, they may reduce their commitment and organisational citizenship behaviours [19]. In the context of supplier-distributor relationship, psychological contract violation has been found to similarly reduce the distributor's trust [24]. Palvou and Gefen (2005) studied psychological contract violation in the buyer-seller relationship in the online marketplace context and found that it has a negative impact on buyers' trust, perceived risk, price premiums, transaction intentions and institutional structures [25]. Furthermore, psychological contract violation has an indirect effect on the customers' intention to reuse online retailer websites through the mediating role of trust and satisfaction [23].

Customer Citizenship Behaviours
In the sharing economy, customer citizenship behaviours are playing increasingly important roles in the development of companies and brands. Customers describe their usage experience online, provide suggestions to the brand, and even participate in new product development. Customer citizenship behaviours, also called extra-role behaviours, refer to discretionary behaviours that are not directly required for the successful production or delivery of the service, but help the organisation overall [2]. Most citizenship behaviour studies have focused on employees or employers [26].
Customer citizenship behaviours are less studied but have attracted notice in recent years. Groth (2005) proposed that customer citizenship behaviours have three dimensions: recommendation, helping other customers and providing feedback. Recommendation occurs when a customer recommends the brand to other customers, such as co-workers, family, and peers. Recommendation is voluntary. It is an indicator of brand loyalty and of great value for brand reputation, since it can attract new customers to the company [2]. Helping other customers includes helping others find products, helping others with their shopping and teaching someone how to use a service correctly. Similar to the importance of voice in organisations [27], feedback from customers is critical for companies to survive in a world of rapid change. Because firms can benefit significantly from customer feedback, a growing number of firms are emphasising the importance of customer feedback. Customers know the details of the product or service. They have considerable experience with the product or service and may have specific needs that are representative of those of a particular market segment. Feedback includes information provided by customers to the firm, and it can help the firm improve its service or product [2]. Customer citizenship behaviours benefit organisations and support a sustainable customer-brand relationship. Previous research has shown that brand reputation [11] and customer commitment to the service worker [10] are strong antecedents of customer citizenship behaviours.

The Effect of Relational Psychological Contract Breaches on Customer Citizenship Behaviours
Studies on employee-organisation relationship have indicated that psychological contract breaches negatively influence employee behaviours, decrease employee trust in their organisation [17], decrease employee commitment to the organisation [28] and decrease voice behaviours and organisation loyalty [29].
In customer-company relationships, psychological contact breaches can have similar negative outcomes. Consumer psychological contract breaches fall into two categories. The relational psychological contract focuses on long-term and interactive, affective relationships, which include not only economic reward but also high affective involvement [18]. Customers may be highly involved with the brand and have many expectations from it. The relational psychological contract inspires a sense of participation and belonging [17]. Because consumers are highly involved in relational contracts, they are more deeply attached to the brand than those in transactional contracts [30], and they are more likely to engage in citizenship behaviours [10]. Fulfilment of the relational contract has a positive effect on organisational citizenship behaviour [19]. Vice versa, breach of relational contract may decrease citizenship behaviours, such as recommending the brand to other consumers and helping other customers. As for the voice behaviours, research from employee-organization literature found that employees respond to psychological contract violation by voicing complaints to their organisations, particularly when employees have a positive working relationship with their supervisors [29]. Consumers who have a relational contract with brands are highly involved with these brands. Therefore, a relational psychological contract breach could increase voice behaviours. Thus, we formed the following hypotheses:

Hypothesis 1c (H1c). A relational psychological contract breach positively affects customer voice behaviours.
When the brand cannot fulfil the relational contract, customers may feel extremely angry. Research has found that psychological contract breaches have a strong relationship with psychological contract violation [17]. We propose that in the customer-brand relationship context, a relational psychological contract breach (RPCB) is positively related to psychological contract violation.

Hypothesis 2 (H2).
A relational psychological contract breach is positively related to psychological contract violation.
Furthermore, psychological contract violation negatively influences consumer purchase behaviours through consumer satisfaction and trust [23]. Consumer satisfaction is positively related to recommendation and helping behaviours [13]. Thus, psychological contract violation may be negatively related to recommendation and helping behaviour. Similar to the relation between psychological contract breach and voice behaviour [29], psychological contract violation may also be positively related to voice behaviour. Consumers increase complaints and suggestions to the companies when they feel psychological contract violation. Therefore, we hypothesise the following:
The conceptual model of the relational contract breach is illustrated in Figure 1.
companies when they feel psychological contract violation. Therefore, we hypothesise the following:
The conceptual model of the relational contract breach is illustrated in Figure 1.

The Effect of Transactional Psychological Contract Breaches on Customer Citizenship Behaviours
Transactional psychological contracts are built on short-term interest and rewards and mainly concentrated on specific and interactive economic relationships [17]. Consumers have low emotional involvement in the transactional psychological contract and have relatively little interaction with the brands [18]. Transactional psychological contract breaches reflect a subjective feeling resulting from others' failure to achieve a functional benefit that is specific, short-term and economic [17]. For example, some retailers use promotions to attract consumers, but they raise the original price of a product before cutting it. Once consumers become aware that the retailer has not given a real discount, they may perceive a transactional psychological contract breach.
Compared with the relational contract, individuals are much less affectively involved in the transactional psychological contract and have much less interaction with the company [18]. Individuals show a lower level of commitment, and engage in fewer reciprocal and interactive behaviours in transactional psychological contracts [30]. We thus propose that transactional contract breaches may directly decrease recommendation and helping behaviours but may not affect the behaviours of voice. Therefore, we hypothesise the following:

Hypothesis 4c (H4c). A transactional psychological contract breach does not affect customer voice behaviours.
Past research has found that psychological contract breaches have a strong relationship with psychological contract violation [17]. Similarly, we propose that in the customer-brand relationship context, a transactional psychological contract breach is also positively related to psychological contract violation.

Hypothesis 5 (H5).
A transactional psychological contract breach is positively related to psychological contract violation.
Psychological contract violation negatively influences consumer purchase behaviours through consumer satisfaction and trust, which are positively related with recommendation and helping behaviours [23]. In a transactional contract breach context, psychological contract violation may also be negatively related with recommendation and helping behaviour. As consumers are less affectively involved in transactional psychological contracts and have less interaction with the company [18] than those of relational contract, we propose that transactional contract breach may not affect voice behaviour directly and indirectly. Therefore, we hypothesise following: Hypothesis 6a (H6a). Psychological contract violation negatively affects customer recommendation; Hypothesis 6b (H6b). Psychological contract violation negatively affects customer helping behaviours; Hypothesis 6c (H6c). Psychological contract violation does not affect customer voice behaviours.
The conceptual model of transactional contract breach is illustrated in Figure 2.

Design and Samples
A 2 (contract type: transactional vs. relational) × 3 (industry: retail vs. catering vs. tourism) between-subject experiment was conducted to test the hypotheses. A total of 601 undergraduate students were recruited and randomly assigned to each level. The subjects were from a well-known university in South China. The respondents who completed the questionnaire were rewarded with 5 credits. The proportion of male participants was 31.8%, and their ages ranged from 18 to 37 (Mage = 22.26, SD = 3.05).

Materials
A pre-test was conducted to select appropriate and valid scenarios for the experiment. We developed two types of psychological contract breach scenarios (transactional and relational). Each kind of breach scenario had three service contexts (retail, tourism and catering). These three contexts represent the most common contexts for participants. According to the definition, a transactional contract is built on short-term interest and mainly concentrated on an interactive economic relationship. Stores use price discounts to attract customers, but they may raise the price before the promotions. Thus, the study used a fake price promotion as a transactional contract breach (Appendix A). Because the relational psychological contract involves a long-term and interactive, affective relationship, we used unfair treatment from sales staff as a relational contract breach. Other elements were the same between the relational contract and transactional contract breach scenarios. We generated six scenarios: 2 (contract types: transactional vs. relational) × 3 (industries: retail vs. catering vs. tourism). The dependent variables in the pre-test were consumer perception of relational and transactional psychological contract breach.
Sixty participants completed the pre-test. Thirty participants completed the transactional psychological contract breach questionnaires, which included three service contexts, and 30 participants completed the relational psychological contract breach questionnaires, which also included three service contexts. As shown in Table 1, for the transactional psychological contract

Design and Samples
A 2 (contract type: transactional vs. relational) × 3 (industry: retail vs. catering vs. tourism) between-subject experiment was conducted to test the hypotheses. A total of 601 undergraduate students were recruited and randomly assigned to each level. The subjects were from a well-known university in South China. The respondents who completed the questionnaire were rewarded with 5 credits. The proportion of male participants was 31.8%, and their ages ranged from 18 to 37 (M age = 22.26, SD = 3.05).

Materials
A pre-test was conducted to select appropriate and valid scenarios for the experiment. We developed two types of psychological contract breach scenarios (transactional and relational). Each kind of breach scenario had three service contexts (retail, tourism and catering). These three contexts represent the most common contexts for participants. According to the definition, a transactional contract is built on short-term interest and mainly concentrated on an interactive economic relationship. Stores use price discounts to attract customers, but they may raise the price before the promotions. Thus, the study used a fake price promotion as a transactional contract breach (Appendix A). Because the relational psychological contract involves a long-term and interactive, affective relationship, we used unfair treatment from sales staff as a relational contract breach. Other elements were the same between the relational contract and transactional contract breach scenarios. We generated six scenarios: 2 (contract types: transactional vs. relational) × 3 (industries: retail vs. catering vs. tourism). The dependent variables in the pre-test were consumer perception of relational and transactional psychological contract breach.
Sixty participants completed the pre-test. Thirty participants completed the transactional psychological contract breach questionnaires, which included three service contexts, and 30 participants completed the relational psychological contract breach questionnaires, which also included three service contexts. As shown in Table 1, for the transactional psychological contract breach scenarios, transactional psychological contract breach scores (mean = 1.8) were lower than the relational psychological contract breach scores (mean = 2.41). In contrast, in the relational psychological contract breach scenarios, relational psychological contract breach scores (mean = 1.81) were lower than transactional psychological contract breach scores (mean = 2.13). One-way ANOVA with the industry factor as an independent variable revealed no significant difference in either the consumer-perceived transactional psychological contract breach (F (2,57) = 0.61, p > 0.05) or the consumer-perceived relational psychological contract breach (F (2,57) = 1.36, p > 0.05). This result indicates that the industry did not significantly affect the dependent variables. As a result, the scenario materials were deemed suitable for the following study since they could prime the corresponding psychological contract breaches.

Experimental Procedure
The experiment had two phases: pre-breach and breach. In the pre-breach phase, the customer-company relationship was favourable. Participants read scenarios about their positive relationship with the brand and reported their level of trust in the brand. In phase 1, we only measured responses to some simple questions, such as trust in the brand, so that participants were more likely to believe the study scenarios. The breach phase was introduced one week later. Participants read a scenario in which the brand did not fulfil its promises. Participants later reported their perception of the contract breach, their perception of psychological contract violation and their customer citizenship behaviours. All the important dependent variables were measured in phase 2. This is a semi-longitudinal design, and we intended to show the progression of a natural relationship (from a good relationship to aversion).

Measures
There are two ways to measure contract breaches. One explicitly assesses contract fulfilment, and the other assesses the discrepancy between obligations and fulfilment. Coyle-Shapiro and Kessler found that although both types of measurement are effective, the first is more methodologically appropriate [19]. We used the explicit method in this study (measurement items are shown in Appendix B). Participants answered the questions using a 5-point Likert scale. The obligations of a relational contract included three items: a long-term warranty, a good service attitude and caring about the customer's feelings. Transactional contract obligations also contained three items: value for money, fair price and many promotions.
The measurement of psychological contract violation was adapted from Eckerd (2013) [31]. The participants were asked to answer the questions using a 7-point Likert scale ranging from "strongly disagree" to "strongly agree". Three items were included to measure psychological contract violation ("I feel betrayed by the brand"; "I feel that my brand has violated the contract between us"; "I feel a great deal of anger towards my brand").
We used the Customer Citizenship Behaviour Scale from Groth [2]. This scale is composed of three dimensions, each with three items: recommendation behaviours (recommending the brand to family, recommending the brand to friends, recommending the brand to people interested in the brand), helping other customers (helping someone use the service correctly, helping others with their shopping, explaining to other customers how to use the service correctly) and customer voice (filling out a customer survey, providing helpful feedback to improve customer service, providing suggestions to better meet my needs).

Reliability
According to the latest methodological literature (Taber, 2018), a good Cronbach's alpha value is typically considered to be above 0.7. However, some other researchers have suggested that a value of alpha above 0.6 is acceptable in social sciences. Cronbach's alpha of all the constructs was above 0.60, which suggests that the internal consistency reliability of each construct was acceptable [32]. All of the reliability data are shown in Table 2.

The Effect of a Relational Psychological Contract Breach on Customer Citizenship Behaviours
The software Mplus 6.0 was used to analyse the data, and SEM was used to test the hypotheses. A total of 302 participants completed the relational psychological contract breach study. In the relational contract breach scenario, the results show an acceptable overall fit: χ 2 = 93.35, df = 80, root-mean-square error of approximation (RMSEA) = 0.024, comparative fit index (CFI) = 0.993 and Tucker-Lewis index (TLI) = 0.99. The SRMR (standardised root mean square residual) was = 0.036 (the goodness-of-fit criteria for 'excellent fit' are: RMSEA ≤ 0.06, CFI ≥ 0.95, TLI ≥ 0.95, SRMR ≤ 0.10 [33]). The correlations of the constructs are shown in Table 3. The results of path analysis show that the relational psychological contract breach had a direct significant negative effect on recommendation behaviours (β = -0.19, p < 0.05), but it had no significant direct effect on voice behaviours (β = 0.08, p > 0.05) or helping behaviours (β = -0.02, p > 0.05). Therefore, H1a is supported, but H1b and H1c are not supported. Furthermore, the relationship between relational psychological contract breach and psychological contract violation was significant and positive (β = 0.51, p < 0.01). Therefore, H2 is supported. Psychological contract violation was found to have a significant negative effect on recommendation behaviours (β = -0.18, p < 0.05) and helping behaviours (β = -0.15, p = 0.10), but it had no significant effect on voice behaviours (β = -0.27, p < 0.001). Therefore, H3a and H3b are supported, but H3c is not supported. Path coefficients are shown in Table 4 and Figure 3.

The Effect of a Transactional Contract Breach on Customer Citizenship Behaviours
The software Mplus 6.0 was used to analyse the data, and SEM was used to test the hypotheses. A total of 292 participants completed the relational contract breach scenario. The results show an acceptable overall fit: χ 2 = 120.83, df = 80, root-mean-square error of approximation (RMSEA) = 0.042, comparative fit index (CFI) = 0.97 and Tucker-Lewis index (TLI) = 0.96. The SRMR (standardised root mean square residual) was = 0.049 (the goodness-of-fit criteria for an 'excellent fit' are RMSEA ≤ 0.06, CFI ≥ 0.95, TLI ≥ 0.95, SRMR ≤ 0.10 [33]). The correlations of the constructs are shown in Table 5.  The results of path analysis show that the transactional contract breach had a direct significant negative effect on recommendation (β = -0.26, p < 0.01) and helping behaviours (β = -0.28, p < 0.01); however, it had no direct significant relationship with voice (β = −0.12, p > 0.05). Thus, H4a and H4b

The Effect of a Transactional Contract Breach on Customer Citizenship Behaviours
The software Mplus 6.0 was used to analyse the data, and SEM was used to test the hypotheses. A total of 292 participants completed the relational contract breach scenario. The results show an acceptable overall fit: χ 2 = 120.83, df = 80, root-mean-square error of approximation (RMSEA) = 0.042, comparative fit index (CFI) = 0.97 and Tucker-Lewis index (TLI) = 0.96. The SRMR (standardised root mean square residual) was = 0.049 (the goodness-of-fit criteria for an 'excellent fit' are RMSEA ≤ 0.06, CFI ≥ 0.95, TLI ≥ 0.95, SRMR ≤ 0.10 [33]). The correlations of the constructs are shown in Table 5.  The results of path analysis show that the transactional contract breach had a direct significant negative effect on recommendation (β = -0.26, p < 0.01) and helping behaviours (β = -0.28, p < 0.01); however, it had no direct significant relationship with voice (β = −0.12, p > 0.05). Thus, H4a and H4b are supported, while H4c is not supported. Furthermore, the relationship between the transactional contract breach and psychological contract violation was significant and positive (β = 0.54, p < 0.01); therefore, H5 is supported. Psychological contract violation had a significant negative effect on recommendation behaviours (β = -0.25, p < 0.01); thus, H6a is supported. However, it had a positive effect on helping behaviours (β = 0.22, p < 0.05); therefore, H6b is not supported. Psychological contract violation had no significant effect on voice behaviours (β = 0.14, p > 0.05); thus, H6c is supported. The path coefficients are shown in Table 6 and Figure 4. contract breach and psychological contract violation was significant and positive (β = 0.54, p < 0.01); therefore, H5 is supported. Psychological contract violation had a significant negative effect on recommendation behaviours (β = -0.25, p < 0.01); thus, H6a is supported. However, it had a positive effect on helping behaviours (β = 0.22, p < 0.05); therefore, H6b is not supported. Psychological contract violation had no significant effect on voice behaviours (β = 0.14, p > 0.05); thus, H6c is supported. The path coefficients are shown in Table 6 and Figure 4. Table 6. Results of TPCB on customer citizenship behaviours.

Discussion
In this study, we tested a framework of psychological contract breaches and citizenship behaviours in the customer-brand relationship context. We assessed the relative impact of two types of contract breaches on customers' recommendation, voice and helping behaviours and the mediating role of psychological contract violation.
The results show that the effect of a relational contract breach was different between dimensions of customer citizenship behaviours: it had a direct significant negative effect on recommendation behaviours, but no direct effect on voice and helping behaviours. We did observe a direct positive effect of relational contract breaches on voice behaviours. It is possible that customers want to express their negative experiences and provide suggestions to the brand. The effect of the relational contract breach on recommendation and helping behaviours was mediated by participants' perception of psychological contract violation. The results indicate that relational contract breach led to a negative emotional experience, and these emotional experiences decreased recommendation and helping behaviours overall. Our findings are generally consistent with the findings of Morrison and Robinson (1997). The relational psychological contract inspires a greater sense of participation and belonging, and customers are highly affectively involved in relational contracts. When a company cannot fulfil

Discussion
In this study, we tested a framework of psychological contract breaches and citizenship behaviours in the customer-brand relationship context. We assessed the relative impact of two types of contract breaches on customers' recommendation, voice and helping behaviours and the mediating role of psychological contract violation.
The results show that the effect of a relational contract breach was different between dimensions of customer citizenship behaviours: it had a direct significant negative effect on recommendation behaviours, but no direct effect on voice and helping behaviours. We did observe a direct positive effect of relational contract breaches on voice behaviours. It is possible that customers want to express their negative experiences and provide suggestions to the brand. The effect of the relational contract breach on recommendation and helping behaviours was mediated by participants' perception of psychological contract violation. The results indicate that relational contract breach led to a negative emotional experience, and these emotional experiences decreased recommendation and helping behaviours overall. Our findings are generally consistent with the findings of Morrison and Robinson (1997). The relational psychological contract inspires a greater sense of participation and belonging, and customers are highly affectively involved in relational contracts. When a company cannot fulfil its relational contract, customers experienced high levels of psychological contract violation, followed by decreased recommendation and helping behaviours.
Furthermore, the results show a very different pattern for the transactional contract breach and customer citizenship behaviours. We found that the transactional contract breach had direct negative influences on two dimensions of citizenship behaviours, namely recommendation and helping behaviours. This finding is also consistent with observations in an organisation context [31]; minor and early contract breaches were shown to potentially trigger negative behaviour directly. However, the transactional contract breach directly led to lower helping behaviours, while transactional contract violation lead to more helping behaviours. This may be attributed to the transactional contract obligations between customers and brands. Customers who have a transactional contract with a brand focuse on specific economic expectations; they may want to help other consumers avoid bad situations after their own violation experiences.
This research makes three theoretical contributions. First, we established the relationship between psychological contract breaches and citizenship behaviours in the customer-brand context. We also extended the literature on the outcomes of psychological contract breaches. Most psychological contract research has focused on transactional contracts in the employee-company context, and research on relational psychological contracts in the customer-company context is nascent [23]. Most of these studies have mainly focused on the effects of psychological contract breaches on the customer's reuse intentions, trust and satisfaction [24]. Our results reveal that the two types of contract breaches have different influences on customer citizenship behaviours. Second, we extended the research scope of customer citizenship behaviours. Customers' active participation plays a more important role than ever before. We uncovered an important antecedent that negatively influences customer citizenship behaviours. Third, we differentiated the role of psychological contract violation into two kinds of relationship. Our findings suggest that the paths by which breaches affect customer citizenship behaviours are different. The results show that psychological contract violation partially mediated the effect of a transactional contract breach on customer recommendation behaviours, and it fully mediated the effect of a relational contract breach on helping behaviours.
The findings above have several managerial implications. Throughout the experiments in our research, transactional contract breaches and relational contract breaches had different effects on customer citizenship behaviours. Therefore, in the context of a contract breach, managers should first identify the type of contract breach. A relational contract breach involves failing to respect the customer's identity or his or her long-term symbolic value, while a transactional contract breach reflects the consumer's disappointment with short-term economic value, such as an unexpected high price or low product quality. Managers should have some strategic awareness that there exist two types of contract breaches that can negatively influence customer citizenship behaviours. In addition, managers should offer convenient channels for customers to provide feedback and suggestions (voice behaviours) to the company. The findings of this research suggest that customers may even increase their voice behaviour when they perceive a relational contract breach. For example, companies can offer connected interactive channels, such as an email address and social media contact, to customers. Moreover, managers should monitor consumers' word of mouth (WOM), since both relational and transactional contract breaches had direct negative influence on consumers' recommendations. If managers can resolve these contract breaches as soon as possible, consumers may not decrease their recommendation behaviours. In other words, in order to maintain a sustainable consumer-brand relationship, managers should focus on the early steps of a contract breach, offer convenient interactive channels and monitor customers' WOM.
This study has some limitations. First, our participants are college students, so the results may have limited external validity. Although college students are suitable for testing theory [34], further investigation is needed to clarify whether other groups exhibit the same patterns. Second, because this study focused on the main effect of contract breaches on citizenship behaviour in the customer-brand context and on the mediating role of contract violation, we did not consider other interpretation processes between the contract breach and contract violation [17]. Since an experimental method was used in our study, the scenarios were hypothetical, and further tests are needed to determine whether other scenarios in other industries produce similar results. At the same time, important factors in the interpretation process, such as fairness judgment and the social contract, were balanced with the hypothetical scenarios. Future studies could explore how the interpretation process moderates the effect of contract breaches on violation in the customer-brand context.

Conclusions
Our experiment assessed the effect of two types of psychological contract breaches on citizenship behaviours in the customer-brand relationship context. The negative effect of a relational contract breach differs between dimensions of citizenship behaviours. A relational contract breach has direct and indirect negative effects on recommendation behaviours and helping behaviours but not on voice behaviours. Customers who perceive a relational contract breach may even increase their voice behaviours. A transactional contract breach directly and negatively influences recommendation behaviours and it indirectly increases helping behaviours through psychological contract violation. Psychological contract violation partially mediates the effect of a transactional contract breach on customer recommendation behaviours and fully mediates the effect of a relational contract breach on helping behaviours.
We established a theoretical relationship between psychological contract breaches and citizenship behaviours in the customer-brand context. We also extended the literature on the outcomes of the psychological contract breaches and expanded the research scope of customer citizenship behaviours. From a practical perspective, the above findings suggest that marketing managers should focus on the early steps of a contract breach, offer convenient interactive channels to increase voice behaviours and monitor customers' WOM in order to maintain a sustainable consumer-brand relationship.

Conflicts of Interest:
The authors declare no conflict of interest.

Appendix A. Scenario of Transactional Psychological Contract Breach
"One afternoon, I passed BAY Supermarket near my house and noticed that there was a sign board in front of the supermarket that read 'In celebration of the third anniversary of BAY Supermarket, 50% sales for all the products!' I wanted to buy a bag of rice. I found that the product had a new price tag: 'Original Price ¥50, Today ¥25'. I remembered the price of this product was ¥ 20 two days before."

Appendix B. List of Measurement Items
Appendix B.1. Relational Contract Breach (1) Brand A offers a long-term warranty.
(2) Brand A offers good service.
(3) Brand A cares about the customer's feelings.