Research on Purchase Intention of E-Commerce Poverty Alleviation Products Based on Perceived Justice Perspective

Abstract

This study is to empirically test whether and how perceived justice affects consumers’ purchase intention regarding e-commerce poverty alleviation products (ECPAP). Based on social exchange theory, self-regulation attitude theory, and the analysis of other relevant literature, this study proposes a model of perceived justice affecting purchase intention regarding ECPAP, in which emotional attitude is the mediator and consumers’ perceived self-serving motivation (CPSSM) is the moderating variable. Three-hundred and eighteen valid observations were collected using the questionnaire method, and these data were analyzed using SPSS 24.0 and AMOS 22.0. The results showed that all three dimensions of perceived justice significantly and positively affect the purchase intention of ECPAP, and this effect is negatively moderated by CPSSM. In addition, an emotional attitude partially mediates the linkage between perceived justice and the intention to purchase ECPAP. The findings not only enrich the research on perceived justice and e-commerce poverty alleviation, but also provide management implications for promoting the benign sales of ECPAP from the perspective of merchants and related subjects. To a certain extent, the above findings can also be extended to other countries’ poverty reduction and rural construction work.

1. Introduction

In all countries worldwide, anti-poverty is a key topic and an important aim of sustainable development. China has listed e-commerce poverty alleviation as one of the ten major works for precise poverty alleviation, with the aim of tackling the problem more efficiently. E-commerce poverty alleviation promotes a seamless connection between poverty-stricken areas’ products and markets, facilitates diversified employment and a convenient income increase for the poor, and realizes the revitalization of rural industries [1,2]. Relevant data show that in 2020, the retail sales for agricultural products in Chinese poor counties reached CNY 40.66 billion, an increase of 43.5% year-on-year [3]. As of August 2021, e-commerce poverty alleviation has driven a total of 7.71 million farmers in poor areas to employment and entrepreneurship and promoted 6.188 million poor people to increase their income [4].

However, unsatisfactorily, e-commerce poverty alleviation faces a real challenge from insufficient growth sustainability. Currently, this approach to poverty alleviation relies mainly on collective procurement or project leadership by the government or institutional units, which makes it hard for poor households to raise their income independently and steadily and makes it easy for them to return to a state of poverty after previously escaping it [5]. To promote the sustainable development of poor areas and the revitalization of villages, we cannot rely solely on “administrative apportionment” or “consumer apportionment.” It is critical to stimulate the market’s effective demand. According to a survey, only around 12% of consumers have actively purchased poverty alleviation products [6]. Meanwhile, in 2020, the electric retail sales of agriculture products in Chinese national poor counties accounted for only 0.15% of all e-commerce commodity transactions [7]. Consumers’ purchase intention for e-commerce poverty alleviation products (ECPAP) is still in need of improvement. The positive degree of purchase intention reflects the degree of consumers’ recognition of ECPAP, which is directly related to the effective degree of consumers’ demand for ECPAP. For this reason, it is important to think deeply about consumers’ purchase intention for ECPAP in order to improve the long-term mechanism of rural construction.

In the research field of poverty alleviation e-commerce, a number of studies have investigated consumers’ purchase intention and behavior toward ECPAP. These studies are based on the SOR theoretical framework, believing that external stimuli (S), such as product characteristics (quality, characteristics, or poverty alleviation labels), online shopping platform factors (regulation of transactions), and social network factors (subjective norms), affect consumers’ trust and behavioral attitude (O), and ultimately influence consumers’ purchase intention (R) for ECPAP [8,9,10,11]. They are more based on a static perspective to study the psychological and behavioral responses made by consumers to environmental stimuli, ignoring the impact of the dynamic process of consumers’ interaction with merchants and other consumers through online platforms. It has been noted that consumers would develop a perception of justice through virtual network interactions with merchants and other consumers [12]. As the basis of social exchange, justice is considered a prerequisite for generating behavioral intention [13]. Unlike direct donations, e-commerce poverty alleviation is not only a moral act, but also a market economy act of “you sell, I buy” [10]. Only by following fair and mutually beneficial market guidelines can we increase consumers’ intention to purchase ECPAP and facilitate the continuous growth of the electric business for poverty alleviation [14,15]. Some studies have analyzed the influence of perceptual factors on individual behavioral intention in the electronic network environment [16]. However, the intrinsic mechanisms and boundary conditions of consumers’ perceived justice affecting the purchase intention regarding ECPAP are unclear, and an empirical study is urgently needed.

In the marketing of products with ethical attributes, consumers’ emotional attitude is often seen as a powerful driver of purchase intention [17,18,19]. Emotional attitude, as an expression of an individual’s overall emotion toward engaging in a behavior [20], can influence an individual’s psychological response to the relevant products [21]. Previous studies have confirmed the positive role of emotional attitude in consumers’ purchase intention in e-commerce contexts. Bilal et al. [22] argued that emotional attitude, as a consumer’s emotional judgment of a specific object, affects consumers’ intention to use WeChat for online purchases. Yu and Zhang [23] showed that emotional attitude is a pre-variable which influences consumers’ intention to purchase products during public-interest live streaming. Zhang [10] also found that consumers’ attitude toward certain behavior is the main predictor of purchase intention in e-commerce poverty alleviation, noting that attitude is influenced by consumers’ perceived factors. Several studies have demonstrated that perceived justice can have a significant impact on consumers’ emotional judgments [24,25,26]. These findings imply that an emotional attitude may work as a mediator in the impact of perceived justice over consumers’ intention to purchase ECPAP. Thus far, no studies have integrated these three factors and clarified their specific relationships.

At the same time, consumers attribute corporate-related activities and motivation when they engage in ethical behaviors [27]. Previous studies have shown that individuals’ perceptions of corporate behavior in fulfilling social responsibility can be divided into two types of motivation: self-serving and other-serving [28,29]. Due to their fundamental nature as commercial entities, consumers may assume that companies are involved in socially responsible activities in hope of economic gain [30]. Many studies have shown that the perceived self-serving motivation of firms affects consumers’ responses to CSR-related activities [28,29]. Among them, some scholars argue that perceived self-serving motivation affects consumers’ purchase intention [31]. Other scholars have examined the moderating role of perceived self-serving motivation in the formation of consumer satisfaction regarding green hotels [32]. It is thus hypothesized that in e-commerce poverty alleviation, perceived merchants’ self-serving motivation may act as a moderator in the process by which perceived justice affects consumers’ purchase intention.

Therefore, the study combines social exchange theory and self-regulation attitude theory to empirically analyze the influence mechanism of perceived justice on the purchase intention of ECPAP from the perspective of perceived justice. The main research questions are as follows: First, does perceived justice affect consumers’ purchase intention for ECPAP, and if so, how? Second, does emotional attitude play a mediating role in the linkage between perceived justice and consumers’ intention to purchase ECPAP? Third, does consumers’ perceived self-serving motivation (CPSSM) moderate the effect of perceived justice on consumers’ purchase intention for ECPAP, and if so, how? The theoretical contributions brought by the research are listed below. First, by integrating two theoretical resources—social exchange and self-regulation attitude—a two-path model of perceived justice affecting consumers’ intention to purchase ECPAP is empirically tested. These are the rational path, which directly influences consumers’ purchase intentions, and the emotional path, in which emotional attitude serves as a mediator. Second, the research considers the moderating role of CPSSM in the influence effect of perceived justice. This helps to unveil the black box of consumers’ purchase decisions regarding ECPAP and deepens the exploration of their boundary conditions. The findings not only enrich the research related to e-commerce poverty alleviation, perceived justice, and consumers’ perceived motivation, but also provide a reference for the sales management of ECPAP.

2. Literature Review and Research Hypothesis

2.1. Literature Review

2.1.1. Purchase Intention of ECPAP

ECPAP are poverty alleviation products that are sold through online platforms. According to the Poverty Alleviation Office of the Chinese State Council, poverty alleviation products originate from poor areas in central and western China and can help to develop poverty-stricken areas, among other benefits [33]. Compared with non-poverty alleviation goods in the market, poverty alleviation products are special products that have both private and public benefits [9]. Poverty alleviation products can enable consumers to achieve material pursuits such as high-quality or personalized products while enjoying psychological satisfaction such as self-satisfaction from helping others. From the level of satisfying needs, the private benefit attribute of ECPAP is more basic [34]. It is also the most important factor in enhancing consumers’ evaluation and recognition of ECPAP [34].

For products, consumers’ intention to purchase is crucial [23]. Most scholars believe online and traditional purchase intentions are similar [35]. Online purchase intention means consumers’ tendency to purchase a product or service in an online shopping context, relying on an online platform [36]. Scholars have conducted numerous studies based on various contexts around online purchase intention in terms of consumer factors, platform characteristics, social network influences, and e-commerce anchor attributes. Mostafa and Hannouf [37] argue that consumers’ online purchase intention in relation to apparel is influenced by perceived benefits and attitude. According to Zhou and Tong’s [38] research, emotional trust and perceived value both influence consumers’ purchasing behavior during live streaming e-commerce. Thus, a paradigm from cognition or emotion to purchase intention has been formed. However, previous studies mostly focused on general products, making it difficult to precisely explain consumers’ purchase decisions for ECPAP with specific attributes. The purpose of e-commerce poverty alleviation is to dissolve consumers’ resistance to poverty alleviation through justice-oriented market transactions to improve the quality and efficiency of poverty alleviation and achieve win–win cooperation [39,40]. Therefore, this study aims to comprehend the mechanism by which consumers’ perceived justice in poverty alleviation transactions affects their purchase intention.

2.1.2. Perceived Justice

Perceived justice originates from social exchange theory, in which individuals in an exchange relationship compare input–output ratios to judge the justice of the exchange [41]. Perceived justice is widely found in the fields of organizational behavior, education, judiciary systems, marketing, etc. Huppertz [42] introduced justice into marketing by stating that consumers also make fair inferences about the goods or services they wish to purchase. This study defines “perceived justice” as consumers’ evaluation of the justice of their transactions with merchants [43]. In the process of communicating and interacting with merchants and other consumers, consumers develop a perception of justice about conducting transactions with merchants. According to the mainstream perspective, perceived justice involves three dimensions: distributive justice, procedural justice, and interactional justice [44,45]. This study also follows this dimensional view. Distributive justice refers to the consumer’s perception of justice in terms of the final outcome of the transaction [43]. Consumers perceive justice to be distributive if the outcome of the purchase is equal to their efforts or meets their expectations. Procedural justice refers to the consumer’s perception of justice with respect to the policies, procedures, and metrics used in the generation of the merchant’s decision [43]. The central logic of procedural justice lies in the allocation of control [46]. Factors such as fair regulations or opportunities in the merchant’s sales process directly influence consumers’ perception of procedural justice. Interactional justice refers to the consumers’ perception of how they are received and viewed by the merchant when communicating interpersonally with the merchant [43]. Consumers experience interactional justice when merchants treat them courteously and considerately and give them adequate information or timely explanations. The performance or importance of the dimensions of perceived justice differs in various contexts, so it is highly significant to extend the research context of perceived justice [47].

2.1.3. Social Exchange Theory and Self-Regulation Attitude Theory

Social exchange theory is a classic theory that explains individual resource exchange behavior. The theory suggests that, guided by the principle of justice, human exchange reflects reciprocity, and individuals tend to respond to proactive behavior with more positive reactions [48]. This theory is premised on the “rational person” assumption that individuals will decide whether to exchange based on maximizing their benefits [49]. Studies have shown that social exchange theory can better predict consumers’ psychological and behavioral decisions [50]. For example, Ting and Thurasamy [51] confirmed, based on exchange theory, that perceived quality affects the revisit intention of infrequent customers of a cafe based on perceived value. Similarly, Li et al. [52] found that both perceived personal preference fit and social relevance affect consumer word-of-mouth behavior in omnichannel retail. In e-commerce poverty alleviation, perceived justice represents a reciprocal benefit that drives consumers to reciprocate merchants with more positive purchase intention.

Self-regulation attitude theory is rooted in the cognitive appraisal theory of emotion and is centered on the self-regulation process of “appraisal–emotional response–behavioral intention” [53]. Specifically, individuals often evaluate their experiences, and the results of these evaluations stimulate emotion, which in turn influences their intention. When an individual has a good experience or achieves a goal, he or she feels an “outcome–desire” harmony that leads to a positive emotional response, which leads to a positive behavioral intention [53]. This theory reveals the emotion-mediated mechanism of value transfer, which integrates the influence of cognition and emotion on individual behavioral responses [54]. This concept is popular among many scholars. It has strong explanatory power in terms of the formation process of variables, such as purchase intention of energy-saving products [55], travel agency customer loyalty [56], and cultural tourism behavior intention [57]. Presumably, the stronger the consumer’s perception of justice, the more positively they will perceive the purchase of that merchant’s ECPAP, which in turn will lead to a more positive purchase intention. He and Luo [58], in their study on customer citizenship behavior, pointed out that combining this theory with exchange theory can give a more comprehensive explanation. Therefore, this study also attempts to combine the above two theories to construct a research framework with both rational and emotional mechanisms for the formation of purchase intention for ECPAP.

2.2. Research Hypothesis

2.2.1. Perceived Justice and Purchase Intention

According to social exchange theory, all three dimensions of perceived justice will enhance consumers’ intention to purchase ECPAP [48]. First, when consumers purchase ECPAP that is both public welfare and a commodity, they attach the greatest importance to the value attributes of the products as commodities [9]. If consumers feel that the distribution is fairer, they will believe that the ECPAP they buy has greater benefits and fewer risks, and thus has higher value [59]. The greater the perceived value, the higher the purchase intention [60]. Second, merchants in the online marketing pattern hold the power advantage provided by information asymmetry [61]. Some merchants use this advantage to disguise poverty, sell substandard products, or even engage in price wars to seize control of the market, which seriously harms the interests of consumers and poor households [62]. The high level of procedural justice perceived by consumers means that consumers have a high level of control over transactions. A higher level of control can narrow the power gap between consumers and the merchant and ensure the order and efficiency of consumption poverty alleviation transactions, thus promoting consumers’ trust towards the merchant. The greater the trust degree of consumers, the stronger the purchase intention [63]. Finally, consumers have less information about ECPAP due to its remote origin and lack of publicity [64]. At the same time, consumers are burdened with information gathering due to a combination of time constraints and an abundance of information-gathering methods [65]. Consumers thus prefer to interact directly with the merchant to obtain information and want to be valued and respected by the merchant. Consumers perceive a high level of interactional justice when they are treated with courtesy and consideration and receive timely responses to inquiries [66]. Thus, interactional justice facilitates the satisfaction of consumers’ psychological needs, such as their dignity, being respected by the merchant, and establishing interpersonal connections with the merchant [67]. The satisfaction of needs can lead to a more positive purchase intention [68]. Based on this concept, we put forward the following hypothesis accordingly:

H1 a.

Distributive justice positively affects the intention to purchase ECPAP.

H1 b.

Procedural justice positively affects the intention to purchase ECPAP.

H1 c.

Interactional justice positively affects the intention to purchase ECPAP.

2.2.2. Mediating Role of Emotional Attitude

Emotional attitude is an individual’s overall emotion (degree of like or dislike) towards performing a particular behavior [20]. The relative influence of the three components—cognitive, emotional, and behavioral—that constitute attitudes differs due to the different levels of consumer motivation [69]. In e-commerce poverty alleviation, consumers have a high level of involvement in purchase decisions. On one hand, to understand what poverty alleviation products are, consumers must gather information, such as corresponding sales channels or poverty alleviation labels, and exert greater cognitive effort [6]. Additionally, consumers also need adequate information about the extent to which their purchases can contribute to anti-poverty efforts [70]. From the other side, the standardization of poverty alleviation products is low and the yield rate is also low [62]. Meanwhile, due to the virtual nature of online transactions, consumers perceive higher risks [71]. Therefore, consumers will spend more time comparing the value attributes of ECPAP and non-poverty alleviation products or offline poverty alleviation products to weigh the advantages and disadvantages.

At the high-intervention effect level, consumers’ attitude toward the purchase decision of ECPAP is based on cognitive information processing, which influences behavior through cognitive influence on emotion [69]. First, when purchasing products with ethical attributes, the primary goal of consumers is to protect their interests from harm [72]. Consumers will measure their interests through justice [73]. If the consumer believes the transaction is fair, they are more likely to believe that their personal interests are protected in the poverty alleviation transaction, and thus they are more likely to believe that the transaction will achieve their primary goal of poverty alleviation consumption. In approaching this goal, individuals develop a more positive affective state [74]. It can be inferred that the easier it is for consumers to achieve their poverty alleviation consumption goals in a transaction, the more consumers will prefer to purchase that merchant’s ECPAP. Second, by the theory of planned behavior, the more positive an individual’s behavioral attitude is, the stronger his purchasing intention will be [75]. When studying Chinese consumers’ trust in food labels, Wang et al. [76] found that behavioral attitude significantly and positively influences purchase intention. In addition, some studies have revealed that in service recovery, perceived justice can promote consumers’ emotional loyalty and thus increase their repurchase intention [77]. Therefore, we proposed the hypothesis:

H2 a.

Emotional attitude mediates the linkage between distributive justice and intention to purchase ECPAP.

H2 b.

Emotional attitude mediates the linkage between procedural justice and intention to purchase ECPAP.

H2 c.

Emotional attitude mediates the linkage between interactional justice and intention to purchase ECPAP.

2.2.3. Moderating Role of CPSSM

CPSSM refers to consumers’ self-serving attribution judgment on the fulfillment of social responsibility by enterprises, i.e., consumers believe that enterprises fulfill social responsibility for business interests or due to social pressure [78]. Different consumers will perceive different degrees of self-serving motivation and thus present different degrees of response [79]. According to social judgment theory, it is difficult for consumers to accept ideas different from their position in the purchase decision of ECPAP with high involvement [69]. Individual participation in poverty alleviation projects is usually motivated by altruism [80]. Consumers who perceive that the merchant is motivated by a self-serving motive will find it difficult to accept the merchant’s ideas because they believe that the merchant is not on the same level as they are [6]. In e-commerce poverty alleviation, consumers’ perception of justice increases their intention to purchase ECPAP. At this point, the greater the level of CPSSM, the larger the perceived difference between the merchant’s idea of poverty alleviation and the consumers’ idea of poverty alleviation. Thus, consumers tend to oppose the idea of the merchant, i.e., they reduce their intention to purchase ECPAP. In other words, CPSSM will weaken the relationship between perceived justice and the intention to purchase ECPAP. In view of this, the following hypothesis is proposed:

H3 a.

The higher the degree of CPSSM, the weaker the effect of distributive justice on the intention to purchase ECPAP.

H3 b.

The higher the degree of CPSSM, the weaker the effect of procedural justice on the intention to purchase ECPAP.

H3 c.

The higher the degree of CPSSM, the weaker the effect of interactional justice on the intention to purchase ECPAP.

Basing on the foregoing literature review and analysis, we proposed the overall framework of the study, as shown in Figure 1.

3. Methodology

3.1. Questionnaire Design

The questionnaire contains four parts. The first part is a greeting and description, including the purpose of this survey, a brief description of the ECPAP, and a thank you to the respondents. The second part is the questions used for sample selection and context introduction. The first question is “Have you ever experienced purchasing ECPAP”, which is used to ensure that the sample information comes from the target group. The second is “How long has it been since your most recent or impressive purchase experience with ECPAP”, which is designed to encourage consumers to answer according to their real feelings at that time. The third part is the variable measurement scales, including the perceived justice scale, the emotional attitude scale, the purchase intention scale, and the CPSSM scale. All questions are based on established scales, with appropriate contextual modifications, as shown in Appendix A. The fourth part is the demographic variables questionnaire, which is applied to understand some personal information of the respondents. A seven-point Likert scale was used, with the numbers “1–7” indicating “strongly disagree–strongly agree”. In addition, the data were statistically analyzed by SPSS 24.0 and AMOS 22.0.

3.2. Pre-Investigation

To ensure the reliability and validity of the results, a small-scale pre-survey was conducted in Jiangsu Province, China, before the large-scale survey. One-hundred and fifty questionnaires were distributed, and one-hundred and twenty-seven valid questionnaires were returned. Reliability analysis and exploratory factor analysis on the pre-survey data were performed by SPSS 24.0. The results manifest that the Cronbach’s α values of each variable were 0.856, 0.883, 0.845, 0.892, 0.825, and 0.935, all greater than 0.8, indicating good reliability of the scale. The Kaiser–Meyer–Olkin (KMO) value of the scale was 0.898, which is greater than 0.8, indicating that the scale is well suited for factor analysis. Six factors are parsed out from 22 items, and the cumulative explained variance is 78.026%. The factor loadings for all questions are greater than 0.5, and there is no cross-loading phenomenon, so the scale validity is good. The questionnaire can be formally used in the study.

3.3. Formal Investigation

The urban residents of Shanghai, Beijing, Zhejiang, Jiangsu, and Guangdong in China were selected for this study for the following reasons. First, the per capita disposable income of residents in the above provinces in 2021 is ranked in the top 6 among all 31 provinces in China. Urban residents in these areas have a higher standard of living and are the main consumers of products from less affluent areas. Additionally, e-commerce and logistical transportation in these areas are better developed, which provides a guarantee for the purchase of ECPAP.

The questionnaire was formally distributed using both online and offline channels. The online survey was performed by distributing questionnaires on such social platforms as WeChat, Qzone, and Weibo. These platforms have a high penetration rate among the consumer groups of ECPAP and can more easily obtain a wider range of random samples. A field survey was conducted by distributing questionnaires at locations such as large shopping centers, stations, and residential neighborhoods. Offline research requires face-to-face communication with respondents, which can make them fill out questionnaires more carefully. Diverse collection channels help ensure the representativeness and reliability of the data.

To reduce possible bias and enhance the quality of the questionnaire data, the respondents were strictly screened. Only respondents who had experience purchasing ECPAP were allowed to fill out questionnaires. For the online survey, IP addresses were also controlled to ensure that respondents met the study requirements. The survey process was also strictly controlled. In the survey, measurement items corresponding to the same variable were separated to eliminate proximity effects. At the same time, trap questions that specified that the specified answer had to be selected were set to enhance the respondents’ motivation to answer accurately.

A total of 360 questionnaires were eventually dispersed, and 345 questionnaires were received. After removing the questionnaires of respondents who had no experience purchasing ECPAP, too short or too long answering times, and incorrect responses to trap questions, we finally obtained 318 valid questionnaires (168 online and 150 offline). The effective recovery rate was 88.3%. The demographic characteristics of the sample are listed in

Table 1. Descriptive statistics of the sample.

Demographic Variable	Classification	The Percentage (%)
Gender	Man	49.06
	Woman	50.94
Age (years)	Under 18 years old	4.09
	18–25years old	34.91
	26–30 years old	47.48
	31–40 years old	11.64
	Over 40 years old	1.89
Educational level	Senior high school and below	1.89
	Junior college	16.35
	Undergraduate	50.31
	Postgraduate and above	31.45
Monthly income (CNY)	Under 2000	8.49
	2000–5000	22.64
	5001–8000	38.36
	8001–11,000	24.21
	More than 11,000	6.29

. They are consistent with the relatively balanced gender, youthfulness, and high education level of Chinese online shopping users. The sample is relatively reasonable.

4. Results

4.1. Reliability and Validity Tests

In this study, confirmatory factor analysis was carried out using AMOS 22.0 to further test the reliability and validity of the scale. AMOS, an analysis software for covariance, has significant advantages in validation factor analysis and is actively used in the marketing field [81,82]. The common algorithm for this software is maximum likelihood estimation, which requires a ratio of sample size to observed variables greater than 10 and normally distributed sample data [83]. The effective sample size of the study is 318, the number of observed variables is 22, and the ratio of the two is more than 10. The maximum absolute value of skewness for all observed variables is 1.236, less than 3, and the maximum absolute value of kurtosis is 2.473, less than 10, and the sample data basically conforms to a normal distribution and can be analyzed in the next step [84]. The analysis results show that χ²/df = 1.871, less than 5; root mean square error of approximation (RMSEA) = 0.052, less than 0.080; goodness of fit index (GFI) = 0.906, comparative fit index (CFI) = 0.961, incremental fit index (IFI) = 0.961, Tucker–Lewis index (TLI) = 0.954, all greater than 0.9. All indicators are better than the fit criteria, indicating that the measurement model has a good fit. As seen in

Table 2. Results of reliability and convergent validity tests.

Construct	Dimensions	Loadings	AVE	CR
DJ	DJ1	0.844	0.676	0.862
	DJ2	0.834
	DJ3	0.788
PJ	PJ1	0.803	0.690	0.899
	PJ2	0.824
	PJ3	0.861
	PJ4	0.834
IJ	IJ1	0.865	0.664	0.887
	IJ2	0.787
	IJ3	0.863
	IJ4	0.738
EA	EA1	0.798	0.685	0.867
	EA2	0.864
	EA3	0.820
PI	PI1	0.834	0.667	0.857
	PI2	0.809
	PI3	0.806
CPSSM	CPSSM 1	0.884	0.752	0.938
	CPSSM 2	0.905
	CPSSM 3	0.823
	CPSSM 4	0.915
	CPSSM 5	0.804

Note: DJ, distributive justice; PJ, procedural justice; IJ, interactional justice; EA, emotional attitude; PI, purchase intention; CPSSM, consumers’ perceived self-serving motivation (the same below).

, the standardized factor loading coefficient of each measurement question is in the range of 0.738 to 0.915, meeting the criteria of greater than 0.5 and significant (p < 0.001); the combined reliability (CR) of each variable is greater than 0.8, which meets the criteria of greater than 0.7; and the average variance extraction (AVE) values are also greater than 0.5. This means that the scale has relatively good combined reliability and convergent validity. As shown in

Table 3. Results of the discriminant validity test.

	DJ	PJ	IJ	EA	PI	CPSSM
DJ	0.822
PJ	0.678 ***	0.831
IJ	0.580 ***	0.606 ***	0.815
EA	0.635 ***	0.698 ***	0.575 ***	0.828
PI	0.636 ***	0.669 ***	0.590 ***	0.664 ***	0.816
CPSSM	0.487 ***	0.488 ***	0.461 ***	0.448 ***	0.464 ***	0.867

Note: The data on the diagonal is the square root of the AVE value of every variable. *** p < 0.010.

, the square root of the AVE value of each variable is greater than the correlation coefficients with other factors, indicating that the discriminant validity among the variables is high.

4.2. Common Method Deviation Test

A common method deviation test was performed to further ensure the reliability of the study results. Considering that each questionnaire in this study was completed by the same subject and tested by self-report, there may be a common method deviation among the variables. The study was tested by both the Harman one-way test and by observing the correlations among the variables. First, unrotated factor analysis was performed with all the measured question items together, and multiple factors were extracted. The first one only accounts for 46.2% of the total variance, which is less than 50%, indicating that the common method deviation is still acceptable [85]. Secondly, the correlation coefficients among all other variables excluding the control variables are in the range of 0.448–0.698 (see Table 3), which is less than 0.9, signifying that the common method deviation is not serious [86]. The conclusions of this study are reliable.

4.3. Hypothesis Test

The study verified the proposed hypotheses with hierarchical regression. Prior to regression, a multicollinearity test was performed. The results show that the maximum variance inflation factor (VIF) value is 2.552, which is smaller than 10, indicating that the issue of multicollinearity does not exist and hierarchical regression can be performed.

4.3.1. The Main Effect Test

Model 1 is a baseline model containing only control variables (i.e., gender, age, education level, and monthly income level). The results are displayed in

Table 4. Results of the main and mediating effects tests.

Variable	Purchase Intention			Emotional Attitude
	Model 1	Model 2	Model 3	Model 4	Model 5
Gender	0.112 *	0.069 *	0.065 *	0.058	0.014
Age	−0.033	−0.058	−0.047	−0.014	−0.041
Educational level	−0.075	0.016	0.005	−0.050	0.045
Monthly income	0.154 **	0.112 ***	0.103 **	0.076	0.032
DJ		0.261 ***	0.197 ***		0.247 ***
PJ		0.356 ***	0.245 ***		0.430 ***
IJ		0.220 ***	0.175 ***		0.173 ***
EA			0.258 ***
R2	0.033	0.552	0.582	0.009	0.558
ΔR2	0.033	0.519	0.030	0.009	0.548
ΔF	2.685 **	119.844 ***	21.842 ***	0.739	128.115 ***

Note: * p < 0.100; ** p < 0.050; *** p < 0.010.

. Model 2 adds three dimensions of perceived justice as independent variables to test the association between perceived justice and the purchase intention of ECPAP. The results indicate that distributive justice (β = 0.261, p < 0.010), procedural justice (β = 0.356, p < 0.010), and interactional justice (β = 0.220, p < 0.010) positively affect the intention to purchase ECPAP, and therefore H1 is verified.

4.3.2. The Mediating Effect Test

The mediating role of emotional attitude was tested based on the three-step approach proposed by Baron and Kenny [87]. Model 2 in Table 4 indicates that distributive justice (β = 0.261, p < 0.010), procedural justice (β = 0.356, p < 0.010), and interactional justice (β = 0.220, p < 0.010) positively affect the intention to purchase ECPAP. As shown in Model 5, the effects of three dimensions of perceived justice on emotional attitudes are all significant (β = 0.247, p < 0.010; β = 0.430, p < 0.010; β = 0.173, p < 0.010). Model 3 shows that in the regression of the dependent variable on the three dimensions of the independent variable with the mediating variables, emotional attitude has a significant effect on the intention to purchase ECPAP (β = 0.258, p < 0.010), and the effects of distributive justice, procedural justice, and interactional justice on the intention to purchase ECPAP all remain significant (β = 0.197, p < 0.010; β = 0.245, p < 0.010; β = 0.175, p < 0.010). Additionally, the regression coefficients of the three dimensions of perceived justice become smaller, which shows that emotional attitudes play a partially mediating role among the three dimensions of perceived justice and intention to purchase ECPAP, and H2 is verified.

4.3.3. The Moderating Effect Test

The moderating effect of CPSSM was detected by using a hierarchical regression model with a product term, as suggested by Wen et al [88]. First, the three dimensions of perceived justice and the CPSSM were centralized, and the product terms between each dimension and the CPSSM were constructed separately. Then, model 6 was constructed with the control variables using the three dimensions of perceived justice, and CPSSM as independent variables and purchase intention as the dependent variable. Based on model 6, the product terms of each dimension of perceived justice and CPSSM were added as independent variables to construct models 7, 8, and 9, respectively. As presented in

Table 5. Results of moderating effects tests.

Variable	Purchase Intention
	Model 1	Model 6	Model 7	Model 8	Model 9
Gender	0.112 *	0.069 *	0.061	0.057	0.062
Age	−0.033	−0.063	−0.049	−0.043	−0.045
Educational level	−0.075	0.031	0.019	0.013	0.019
Monthly income	0.154 **	0.123 ***	0.114 ***	0.114 ***	0.113 ***
DJ		0.237 ***	0.225 ***	0.223 ***	0.240 ***
PJ		0.334 ***	0.295 ***	0.278 ***	0.304 ***
IJ		0.199 ***	0.188 ***	0.176 ***	0.171 ***
CPSSM		0.108 **	0.108 **	0.104 **	0.101 **
DJ×CPSSM			−0.100 **
PJ×CPSSM				−0.140 ***
IJ×CPSSM					−0.096 **
R2	0.033	0.560	0.567	0.572	0.566
ΔR2	0.033	0.527	0.007	0.012	0.006
ΔF	2.685 **	92.514 ***	4.812 **	8.595 ***	4.279 **

Note: * p < 0.100; ** p < 0.050; *** p < 0.010.

, the explanatory power of models 7, 8, and 9 is significantly increased by 0.007, 0.012, and 0.006, respectively. The CPSSM negatively moderates the linkage between distributive justice (β = −0.100, p < 0.050), procedural justice (β = −0.140, p < 0.010), interactional justice (β = −0.096, p < 0.050), and intention to purchase ECPAP; therefore, H3 is verified.

4.4. Robustness Test

To strengthen the persuasiveness of the results, this study conducted robustness tests on the hypotheses.

4.4.1. The Main Effect’s Robustness Test

Referring to previous studies [89,90], structural equation modeling operations were used to re-test the significance of the path coefficients and determine the robustness of the main effect. The fit indices of the structural model met the requirements (χ²/df = 1.936, less than 5; RMSEA = 0.054, less than 0.080; CFI = 0.966, TLI = 0.959, GFI = 0.927, IFI = 0.967, all greater than 0.9). The path results are shown in

Table 6. Validation results of the structural equation model.

Path	Estimate	Standardized Estimate	S.E.	C.R.	P
DJ→PI	0.185	0.220	0.090	2.050	**
PJ→PI	0.202	0.237	0.100	2.026	**
IJ→PI	0.148	0.160	0.078	1.891	*
DJ→EA	0.240	0.303	0.079	3.036	***
PJ→EA	0.400	0.498	0.081	4.942	***
IJ→EA	0.124	0.143	0.072	1.721	*
EA→PI	0.364	0.343	0.125	2.899	***

Note: * p < 0.100; ** p < 0.050; *** p < 0.010.

, with each dimension of perceived justice significantly and positively influencing the intention to purchase ECPAP (β = 0.220, p < 0.050; β = 0.237, p < 0.050; β = 0.160, p < 0.100). H1 is again verified, i.e., the main effect is robust. In addition, the other path coefficients are also significant, which provides initial support for retesting the mediating effect.

4.4.2. The Mediating Effect’s Robustness Test

With the aid of the PROCESS plug-in, the Bootstrap method was adopted to test again the mediating effect of emotional attitude. The sample was replicated 5000 times, and a 95% confidence interval was selected. The results can be found in

Table 7. Robustness test results of the mediating effect.

Path	Effect Type	Effect	Lower 95% CI	Upper 95% CI	Zero Included?
DJ→EA→PI	Direct	0.181	0.085	0.277	No
	Indirect	0.059	0.022	0.105	No
PJ→EA→PI	Direct	0.232	0.120	0.344	No
	Indirect	0.111	0.056	0.171	No
IJ→EA→PI	Direct	0.174	0.076	0.273	No
	Indirect	0.047	0.018	0.085	No

. In the impact of perceived justice upon the intention to purchase ECPAP, the confidence intervals corresponding to both direct and indirect effects do not include 0, demonstrating that emotional attitude has a partial mediating role in it. H2 is again verified, i.e., the mediating effect of emotional attitude is robust.

4.4.3. The Moderating Effect’s Robustness Test

Using the method of reducing the control variables to further test the moderating effect [91], if the sign and significance of the product term of the independent variable and the moderating variable do not change significantly after removing the control variables that have a significant effect on the regression results, the moderating effect is robust [91,92]. As presented in Table 5, there is a significant effect of gender and monthly income level on the regression results.

Table 8. Robustness test results of the moderating effect.

Variable	Purchase Intention
	Model 10	Model 11	Model 12	Model 13	Model 14
Age	0.005	−0.029	−0.015	−0.008	−0.011
Educational level	−0.066	0.039	0.025	0.019	0.024
DJ		0.252 ***	0.237 ***	0.236 ***	0.254 ***
PJ		0.338 ***	0.291 ***	0.275 ***	0.302 ***
IJ		0.199 ***	0.186 ***	0.173 ***	0.165 ***
CPSSM		0.092 **	0.094 **	0.089 *	0.086 *
DJ × CPSSM			−0.118 **
PJ × CPSSM				−0.156 ***
IJ × CPSSM					−0.115 **
R2	0.004	0.544	0.554	0.559	0.553
ΔR2	0.004	0.540	0.010	0.015	0.009
ΔF	0.681	92.205 ***	6.626 **	10.513 ***	6.124 **

Note: * p < 0.100; ** p < 0.050; *** p < 0.010.

shows that after removing these two, all product terms remain significantly negative, and H3 is again validated. The moderating effect of the CPSSM is robust.

5. Discussion

This study is conducted to empirically investigate the influence mechanism of perceived justice on consumers’ intention to purchase ECPAP.

First, the findings suggest that perceived justice has a positive impact on consumers’ intention to purchase ECPAP. The conclusion is consistent with the results of previous research on brand perceived justice and customer purchase intention [93]. On the principle of reciprocity in social exchange, when consumers perceive a high degree of justice, they will reward merchants for increasing their performance by purchasing their ECPAP. In terms of the degree of influence of internal dimensions, procedural justice has the strongest predictive power for poverty alleviation in terms of consumers’ purchase intention, followed by distributive justice and interactional justice. This finding differs from those of previous studies in that interactional and distributive justice have a more significant positive effect than procedural justice in online shopping [94]. This may be because, compared with general products, most of the ECPAP are produced in remote mountainous areas with lower standardization, poorer quality, and have a public interest attribute, and the favorable degree of consumer purchase outcomes is relatively small. Additionally, it is only in the face of relatively unfavorable outcomes that procedural justice comes into play, and its impact may be greater as a result [95]. Rich social media resources provide consumers with access to information about a merchant’s sales policies or program operations. In addition, to boost sales and reputation, some merchants use various tactics to lure and deceive poverty alleviation consumers, and the authenticity of online interactions is in doubt [96]. Thus, the impact of distributive justice and interactional justice on the purchase of ECPAP is relatively small.

Second, emotional attitude takes a mediating role in the association between perceived justice and consumers’ purchase intention of ECPAP. This is similar to the findings of previous studies related to the hotel and insurance industries, which found that consumers’ emotional response mediates the association between perceived justice and repurchase intention [24,97]. According to self-regulation attitude theory, the cognitive assessment of justice generated by consumers will stimulate the emergence of an emotional response to the act of purchasing ECPAP, which influences the corresponding purchase intention [53]. This finding signifies the key role of consumers’ emotional attitude in e-commerce for anti-poverty and echoes the findings of Zhang [10], who revealed that consumers’ attitude is the main driver of purchase decisions for ECPAP. This study also clarifies the partial mediating role of emotional attitude between perceived justice and purchase intention of ECPAP. That is, both the direct rational path and the indirect emotional path of perceived justice affecting the purchase intention of ECPAP are verified. This supports the view that poverty alleviation consumer groups blend rational perceptions with emotional empathy [62]. Additionally, the distinction between being a partial mediator vs. a full mediator suggests the necessity for exploring other mediating variables in the future [98].

Finally, CPSSM acts as a negative moderator between perceived justice and consumers’ intention to purchase ECPAP. The finding is in line with prior research. According to Shen and Fan [99], consumer perceptions of motivation play a moderating role in consumer responses induced by the socially responsible behavior of online retail companies. Tian and Xiao’s [78] study of companies engaged in socially responsible activities demonstrated that consumer perceptions of self-serving motivation negatively moderate the link between perceived quality and purchase intention. When it comes to e-commerce poverty alleviation, consumers believe in helping others and expect merchants to maintain a consistent stance [6]. In the process of generating consumers’ purchase intentions for ECPAP based on perceived justice, consumers following the altruistic principle tend to oppose the self-serving ideas of merchants and thus become less motivated to purchase. Thus, the higher the degree of CPSSM, the weaker the effect of perceived justice on consumers’ purchase intention of ECPAP.

6. Theoretical Contributions

Three main theoretical contributions exist in the study.

First, the study compensates for the lack of previous research that has less considered the impact of the dynamic process of pro-poor consumers’ interactions with merchants and other consumers, providing a perceived justice perspective for understanding the purchase intention in the background of e-commerce pro-poor transactions. Although studies have investigated the impact of external stimuli in terms of product, trading platform, and social network on the purchase intention of ECPAP, these studies are only from a static perspective and ignore the important role of merchants. The dual attributes of market and public welfare of ECPAP also distinguish them from general products, and the available studies on the factors influencing the purchase intention of other products have provided limited guidance. The fundamental purpose of e-commerce poverty alleviation is to stimulate the sustained growth of pro-poor consumption through justice-oriented market transactions. This study provides a more focused reference for scholars to understand the purchases of e-commerce pro-poor consumers.

Second, the study confirms the dual-path model of perceived justice influencing purchase intention of ECPAP, i.e., the direct influence path based on rational judgment and the indirect influence path mediated by emotional attitude. Reviewing the literature, we found that academics have studied the association between perceived justice and emotional response, emotional attitude, and purchase intention, respectively, based on various contexts. However, the association between perceived justice, emotional attitude, and intention to purchase under the context of e-commerce poverty alleviation has yet to be considered. This study combines social exchange theory and self-regulation attitude theory to build a more complete framework of perceived justice affecting the purchase intention of ECPAP. This not only enriches the research on e-commerce poverty alleviation and perceived justice, but also extends the application contexts of social exchange theory and self-regulation attitude theory.

Finally, the study reveals the moderating role of CPSSM in the growth of e-commerce poverty alleviation. Among previous scholars’ attempts to explore the purchase intention of ECPAP, the antecedent variables have received more attention, with insufficient attention given to the boundary conditions. Based on the related research on perceived motivation, the introduction of CPSSM as a moderating variable deepens and improves the influence mechanism of perceived justice on the purchase intention of ECPAP, which is enlightening for subsequent research on the moderating factors between the two relationships.

7. Management Implications

The findings of the study offer beneficial management implications for promoting benign sales of ECPAP.

First, consumers should be treated fairly. Merchants should realize that procedural justice is a key factor in promoting poverty alleviation consumption. Fair policies and systems for public benefit sales should be carefully established to ensure that they are reasonable, consistent, and modifiable. In daily sales, procedures must be strictly adhered to and not changed at will. For major decisions involving the interests of pro-poor consumers, merchants can offer consumers the opportunity to be informed, get involved, and express their views by setting up or opening comment sections and private chat functions on online sales platforms. Based on the implementation and feedback of the sales policy, the procedure should be optimized and improved in a timely manner to enhance consumers’ perception of procedural justice. Next, merchants should try to secure a reasonable gain/loss ratio for consumers in order to enhance their perception of distributive justice. To this end, it is necessary to clarify product standards that are in line with the market, establish an information-based traceability platform, and realize quality management and control of the entire chain of production, storage, transportation, and sales of ECPAP. Combined with the consumer demand, we should promote the deep processing of ECPAP as much as possible, enrich the product variety, prolong the industrial chain, and improve the value-added of the products. At the same time, it is also necessary to reasonably price the ECPAP according to market supply and demand, and it is advisable to create regional special product brands based on the characteristics of poor areas to reduce unnecessary payment or identification costs.

In addition, merchants should fully respect and care about consumers’ demands to enhance consumers’ perception of interactional justice. Social media can be utilized to communicate with consumers worldwide and around the clock to achieve real-time responses. When interacting with consumers, merchants ought to maintain a friendly service attitude, present clear and complete information about ECPAP, and avoid deliberately concealing unfavorable information. For consumers with reading difficulties, it is necessary to combine text, sound, and video to present and deliver information about poverty alleviation products.

Consumers’ emotional attitude must also be considered. Consumers’ liking for a merchant’s poverty alleviation products helps sell the products, but perceived justice is not the only factor that contributes to an emotional attitude. Feelings such as the sense of morality and accomplishment that come from helping others in a charitable transaction can also be effective in boosting consumers’ favorability toward a merchant’s products. Therefore, merchants should focus on promoting the public welfare attributes of poverty alleviation products to stimulate consumers’ spirit of helping others without threatening their private interests. For example, merchants can try to integrate the stories of poor households into the packaging or advertising design of ECPAP to mobilize consumers’ feelings of poverty alleviation from multiple standpoints, such as visual and auditory perspectives. At the same time, by inviting consumers to join in the creation of logos for poverty alleviation brands and the conception of corporate poverty alleviation proposals, consumers can be guided to think positively and experience the pleasure of contributing, forming a positive attitude toward buying merchants’ products.

Additionally, merchants should aim to actively fulfill their social responsibilities. The motivation of businesses, as economic entities, to engage in socially responsible activities is susceptible to consumer skepticism. Poverty alleviation consumers’ perception of merchants’ self-serving motivation can undermine the positive effects of perceived justice. Merchants should minimize consumer suspicion of their own self-serving motivation for socially responsible behavior. To do so, they should enhance stakeholder awareness, establish a win–win logic, and view CSR correctly. In the process of fulfilling their responsibilities, they should complete authoritative identification of poverty alleviation qualifications, improve the benefit distribution mechanism for poor households, strictly follow the management standards of trading platforms, and operate in a standardized and sincere manner. They should resolutely eliminate marketing under the name of poverty alleviation and ensure that the profits of poor producers are real and reasonable. If necessary, they should strengthen their cooperation with third-party charitable organizations and let third-party organizations release information on the poverty alleviation of enterprises to enhance the credibility of poverty alleviation actions. In addition, consumers may behave differently in corporate motivation attribution due to different cultural backgrounds. The findings of this study should remind all companies to pay attention to and be cautious about consumers’ perceptions regarding CSR activities, rather than abandoning them.

In addition to the merchants’ actions, stakeholders should also show initiative. Consumers should realize that e-commerce poverty alleviation can not only help to meet their own product needs, but can also help with the development of poor areas and promote social development and solidarity. On this basis, it can raise awareness of poverty alleviation consumption, strengthen the supervision of poverty alleviation merchants, and drive poverty alleviation merchants to operate in good faith and help the poor. The trading platform should strictly review the access of pro-poor merchants to keep unscrupulous merchants from exaggerating poverty alleviation and damaging the consumption atmosphere of the platform. The government should provide preferential policies or fee waivers for poverty alleviation merchants in terms of operation and taxation to enhance their motivation to fulfill their social responsibility. At the same time, relevant laws and ordinances can be improved to strengthen the management of the poverty alleviation trading market and safeguard the interests of merchants, consumers, and poor households.

8. Limitations and Future Research

The study has some limitations and can be further improved in future work. First, the respondents of this study were exclusively Chinese poverty alleviation consumers. The findings should be revalidated in a cross-cultural context to improve the applicability of the model. Second, this study only studied the mediating role of emotional attitude between perceived justice and purchase intention in ECPAP. The consumers’ decision-making process is complex, and other mediating variables can be selected in the future to deepen the study of the influence mechanism. In addition, the study only explored the moderating effect of the CPSSM. The possible existence of other moderating variables, such as consumers’ sensitivity to justice, consumers’ purchasing ability, and time, could be considered in the future to enrich the exploration of the boundaries of the perceived justice effect.

9. Conclusions

Based on social exchange theory and self-regulation attitude theory, the study empirically examines the influence mechanism of perceived justice on consumers’ intention to purchase ECPAP. Specifically, perceived justice can directly promote consumers’ purchase intention in relation to ECPAP, in which the positive effect of procedural justice is particularly prominent. In addition, perceived justice also indirectly enhances purchase intention when mediated by emotional attitude. Moreover, the intensity of the effect of perceived justice on the purchase intention of ECPAP varies depending on the degree of CPSSM. The higher the degree of CPSSM, the weaker the effect of perceived justice on consumers’ intention to purchase ECPAP. Therefore, poverty alleviation merchants should treat consumers fairly, pay attention to their emotional attitude, and actively assume social responsibility. Stakeholders such as consumers, trading platforms, and governments can also contribute.