Why does browsing social networking sites increase conspicuous consumption? The effects of relative deprivation and subjective socioeconomic status

Abstract

Background

Conspicuous consumption is a type of consumption behavior aimed at displaying one’s wealth and status, which may impose an economic burden and negative psychological impact on consumers. With the widespread spread of conspicuous consumption among young people in China, it is of great significance to explore its antecedents. However, previous studies on the association between passive social networking site use and conspicuous consumption remain insufficient.

Purpose

The present research attempted to investigate the relationship between passive social networking site use and conspicuous consumption and its internal mechanisms.

Methods

In Study 1 (pilot study), 138 participants were recruited for a preliminary analysis using a questionnaire method. In Study 2, we recruited 203 participants using a univariate inter-group design (passive social networking site use vs. non-passive social networking site use). In Study 3, we recruited 163 participants using a 2 (passive social networking site use vs. non-passive social networking site use) x 2 (lower subjective socioeconomic status vs. higher subjective socioeconomic status) inter-group design.

Results

The results show that: (1) Passive social networking site use positively predicted conspicuous consumption. (2) Relative deprivation mediated the positive relationship between passive social networking site use and conspicuous consumption. (3) Subjective socioeconomic status directly and indirectly (through the sense of relative deprivation) moderated the relationship between passive social networking site use and conspicuous consumption.

Conclusion

This research extends the relevant findings on consumption behaviors under the Internet scenario. In particular, the research provides new insights into reducing conspicuous consumption in the context of passive social networking site use.

Clinical trial number

Not applicable.

Introduction

From the point of view of the “economic man”, a comprehensive trade-off of factors such as price and quality is the main determinant of the sales potential of products [1]. Nevertheless, the contemporary consumption pattern is largely guided by the non-functional symbolic attributes of brands [2]. Such behaviors of purchasing and exhibiting expensive goods with symbolic attributes to impress others that one has wealth or status are labeled as “conspicuous consumption (CC)” [3]. In recent decades, CC has spread broadly among the Chinese young generation [4]. However, given the limited savings of many young people, CC may adversely affect their financial situation and increase their indebtedness [5]. Moreover, CC also has some unfavorable psychological outcomes. For instance, consuming expensive products that don’t match an individual’s actual identity increases the sense of inauthenticity and thus reduces subjective well-being and confidence [6, 7]. Therefore, it is necessary to identify the antecedent variables of CC among young people.

Previous studies have revealed that certain psychological factors can predict CC, such as social status display [8], the desire for access to elite social strata [9], power deprivation [10], and lower subjective socioeconomic status [11]. What these psychological factors have in common is that they all involve the motivation of seeking status. Considering that conspicuous products are symbols of wealth and status, it is possible that people buy and display these conspicuous products (i.e., CC) in pursuit of status to construct an ideal self-image [3, 8, 12]. Moreover, with the development of the Internet, behaviors intended to present a good image (including CC) are no longer limited to the offline scene, but have begun to appear frequently in the network [12, 13]. In particular, social networking sites (SNSs) provide users with a platform for the creation and exchange of user-generated content [14, 15], which makes it more convenient to display a variety of conspicuous products. As related studies have illustrated, the usage of SNSs can positively predict CC [13, 16, 17, 18, 19]. In addition, the young population is the main user group of SNSs [20], and for them, SNSs are also an integral part of life. Therefore, SNS use may be one of the important environmental variables affecting their CC.

However, while previous studies have generally confirmed the positive predictive effect of SNS use on CC, their measures of SNS use are mostly based on self-report of the intensity of overall usage, without differentiating between specific patterns of use. Generally, individuals can participate in SNSs in two ways. One is active use, which refers to activities that facilitate direct exchanges with others (e.g., direct communication and broadcasting); the other is passive use (PSU), which refers to passively browsing on SNSs and consuming information without direct communication (e.g., viewing other users’ profiles and scrolling through news feeds) [14, 21, 22]. In fact, it seems that the psychological and behavioral consequences of these two patterns of SNS use are inconsistent [14, 15]. PSU may be associated with more negative emotional and cognitive outcomes especially [14, 15, 23]. Nevertheless, most youths participate in SNSs mainly for passively browsing (called passive users) rather than actively posting (called active users) [22], making it meaningful to investigate the relationship and psychological mechanisms between PSU and CC among young people. Furthermore, since the method used by previous studies was mainly based on self-report, which did not allow causal inferences, this research considered using an experimental method to explore the predictive effect of PSU on CC.

Passive social networking site use and conspicuous consumption

Based on the cultivation theory, media constructs users’ perception of social reality and shapes their values about material and consumption [24]. Cultivation effects result from the steady and recurring portrayals in the media [25]. By exposing users to certain information frequently, media spread certain values and norms of behavior (including those related to consumption). Among various media, SNSs provide a platform for the public to generate diverse content and opinions and widely spread them [15, 22]. According to Khamis and colleagues [26], this transmission of information on SNSs is mainly driven by a specific kind of identity construction (self-mediation), aimed at maintaining a good self-image. Therefore, what users post and share online is well-embellished and intended to be shown to others. For image management purposes, users may show off their wealth or social status on SNSs, such as displaying luxury products or expensive services [12, 13]. Therefore, CC-related symbols are embedded in the content shared by SNS users. Due to the increased accessibility of CC-related content while browsing SNSs, PSU users are motivated to imitate CC [24]. As a result, PSU may prompt people to engage in CC. Related research has shown that PSU significantly positively predicts materialism [27] and overspending [19], both of which are closely associated with CC. Thus, we proposed our first hypothesis:

H1

PSU can positively predict CC.

The mediating role of relative deprivation

If PSU can increase CC tendency, what is the internal mechanism? The second goal of this research was to investigate possible mediating factors between PSU and CC and fill in the gap. Based on the compensatory consumption behavior model, the consumption of products is sometimes driven by the motivation to compensate for psychological assets [28]. For conspicuous goods, it is a symbolic signal of wealth and status [3]. Thus, psychological factors associated with perceived inferiority of wealth and status, such as the sense of relative deprivation (RD), may be potential mediating factors.

RD is an individual self-judgment that regards oneself as worse off compared to some standards, accompanied by feelings of anger and resentment due to the discrepancy from the ideal state [29]. RD could arise from PSU. As mentioned earlier, when users are just silently observing others on SNSs, they usually receive lots of positively skewed information presenting a perfect self and an ideal life [26]. Exposing to idealized information of others may motivate upward comparison [15, 23, 27, 30, 31]. Upward comparison is a form of social comparison with others who are better than themselves in some aspects, which is often accompanied by dissatisfaction, perceived unfairness, and jealousy and resentment toward others [15, 30, 31, 32, 33], which lead to the feeling of RD [34]. Therefore, we considered that PSU can positively predict RD.

Furthermore, the sense of RD is likely to evoke unpleasant responses (such as stress) [35]. As a result, people are encouraged to self-regulate in order to improve the current situation [29], and these adjustments can be reflected in consumption behaviors [36]. On the ground of the compensatory consumption behavior model, the motivation to reduce the sense of RD caused by self-discrepancy impels individuals to engage in high-consumption behaviors [28]. Considering ostentatious goods and services represented by articles of luxury are often seen as status symbols, they are capable of helping people symbolically cope with the discrepancy between the ideal self and the real self in power and status [37]. For instance, research has revealed that individuals whose self-worth is threatened or who feel powerless seek high-status goods to treat psychological woods [36, 38]. Herein, we proposed that PSU stimulates the sense of RD due to upward comparison. In turn, passive users engage in CC in order to make up for the unsatisfied psychological needs brought about by RD. Here’s our second hypothesis:

H2

RD mediates the positive relationship between PSU and CC.

The moderating role of subjective socioeconomic status

Based on the differential susceptibility to media effects model [39], which assumes that the influences of media depend on individual variables and social context, we took the moderating effects of subjective socioeconomic status (SSS) into consideration. Socioeconomic status (SES) is a pervasive form of hierarchy that refers to an individual’s rank in society in terms of income, occupation, and educational qualification [40]. In addition to these objectively measured indicators, people also have a perception of where they reside in the social hierarchy (i.e., SSS) [40]. SSS has the advantage of being malleable, making it a useful measure for examining causality [11].

In terms of the non-functional symbolic value of conspicuous goods, individuals of different SES may all engage in CC [41]. Although common intuition is upper-class individuals are more financially able to afford luxury goods, there is evidence that individuals with lower SSS are more likely to engage in CC [11, 42]. Does lower SSS increase the positive predictive effect of PSU on CC? We can discuss this in two aspects.

On the one hand, the sense of scarcity triggers a focus on scarce-related demands [43]. For example, lower-income people pay more attention to money-related cues [44]. It can be assumed that individuals with lower SSS are more inclined to pay attention to relevant content with status symbols when browsing SNSs. This may encourage them to emulate this consumption pattern [24, 27], that is, to engage in CC.

On the other hand, the content of SNSs is glorified. Individuals with lower SSS are likely to focus on content that conveys the message that others are superior to them in terms of wealth and status. This promotes upward comparison [15, 23, 27, 30, 31], which triggers feelings of anger and perceived unfairness, also known as RD [34, 45]. Thus, to compensate for the sense of RD induced during browsing SNSs, individuals with lower SSS are more likely to engage in CC. To sum up, we proposed the following hypothesis:

H3

There is an interaction effect between PSU and SSS. To be specific, in the PSU group, individuals with lower SSS have a higher sense of RD and CC tendency than individuals with higher SSS.

The present research

The main purpose of this research was to explore the effect of PSU on CC and its underlying internal mechanisms. In Study 1 (pilot study), we first conducted a preliminary analysis of the correlation between PSU and CC by self-reported measures. In Study 2 and Study 3, we recruited college students from two regions of China and tested the predictive effect of PSU on CC by manipulating PSU. Moreover, we also respectively investigated the mediating role of RD (Study 2 and Study 3) and the moderating role of SSS (Study 3).

Study 1 (pilot study)

Method

Participants

In this pilot study, a total of 150 questionnaires were distributed in universities in Northwest and Southwest China through a combination of online and offline. 138 valid questionnaires (92% effective recovery rate) were obtained after excluding 12 participants who answered regularly. There were 47 males and 91 females, aged from 18 to 26 (M = 20.03, SD = 1.96). All participants were informed of the purpose and the confidentiality measures.

Measures

Daily passive social networking site use

The Chinese version of Surveillance Use Scale [46] was used to measure the intensity of individual daily PSU. The original scale was compiled by Tandoc and colleagues [47]. The scale consists of four items, including reading the ‘newsfeed’; reading a friend’s status update; viewing a friend’s photo; and browsing a friend’s timeline. In order to avoid confusion between active use and passive use, the scale instruction reminded participants to respond according to the situations when they use social networking sites without posting updates or chatting with others. We employed a 5-point scale to rate, where 1 = never and 5 = frequently. A higher score indicates a greater intensity of PSU. The Cronbach’s α of this scale in this study was 0.78.

Conspicuous consumption

The Chinese version of Conspicuous Consumption Scale translated and revised by Chinese scholars [48] was adopted. The original scale was derived from the study of Marcoux and colleagues [49]. The revised scale consists of eighteen items, including “owning brand-name goods means being rich.”. A 7-point Likert scale was used, ranging from 1 (completely disagree) to 7 (completely agree), with a higher score indicating a higher tendency to CC. The Cronbach’s α of this scale in this study was 0.92.

Statistical analysis

SPSS 26.0 was used to perform Pearson correlation analysis. In all three studies, p < 0.05 indicates a significant level.

Results

Table 1 lists the descriptive statistics (means and standard deviations) and correlation analysis results of each variable. Among them, there is a positive and significant correlation between daily PSU and CC (r = 0.298^**, p < 0.001), but neither of them significantly correlated with age or gender.

Table 1 Descriptive statistics and correlations for variables of study 1 (N = 138)

Study 2

Method

Participants

In this study, 203 college students from Southwest China were recruited. There were 93 males and 110 females, with an age range from 17 to 25 years old (M = 19.69, SD = 1.59). They all volunteered to participate in the study and obtained formal consent.

Measures

Daily passive social networking site use

The measuring tool was consistent with Study 1, and the Cronbach’s α of this scale in this part was 0.78.

Manipulation of passive social networking site use

Reference to the manipulation method of Vogel and colleagues [50] and in combination with the usage of domestic social networking sites, each participant in the PSU group (experimental group) was required to browse previous posts of their followers on Sina Weibo for 5 min and not to interact with friends or publish content (distinguish it from the active use behavior). The non-PSU group (control group) viewed the CCTV news website for 5 min. At the beginning of the experiment, the participants were told to answer relevant questions after the browsing task to ensure serious participation. The manipulation check was carried out using our three additional items, including “Whether the information you have browsed just now was consistent with the social network information (Sina Weibo, WeChat circle of friends, etc.) you browse in daily life?”. A 7-point Likert scale (1 = not at all, 7 = very much) was adopted, with a higher score indicating a higher degree of daily PSU.

Relative deprivation

The Self-compiled Chinese version of Relative Deprivation Questionnaire [51] was adopted, which includes four items such as “My life should be better than it is compared to the effort I have made.”. A 6-point Likert scale was applied, with 1 = strongly disagreement and 6 = strongly agreement. And a higher score means a higher level of RD. The Cronbach’s α of this questionnaire was 0.80.

Conspicuous consumption

Based on Lee and Shrum’s measurement [52], we constructed a hypothetical market survey scenario, informing participants that they were about to engage in a market survey on fashion design. The participants were presented with two hoodies with the same price and quality, the only difference being the size of the logo. And then participants were required to answer two questions (for instance, “If we gave you a hoodie for everyday wear, which one would you choose?”). Our questions were in Chinese which are derived from the study of Xie and colleagues [53]. The questions were scored on a 9-point scale (the closer the score is to 1, the more inclined to the small logo, and the closer the score is to 9, the more inclined to the large logo). A higher score indicates a higher propensity for CC. The Cronbach’s α of the two items in this study was 0.98.

Design and procedure

In this part of the study, a univariate inter-group (PSU vs. non-PSU) design was adopted, with PSU as the manipulated variable. The participants were randomly divided into PSU group and non-PSU group, in which the former group included 103 participants (50 males and 53 females) and the latter group included 100 participants (42 males and 58 females).

The specific process was as follows:

• (1) Participants first filled out demographic information and the Surveillance Use Scale.

• (2) PSU was manipulated and participants were required to answer related questions after manipulation.

• (3) Participants finished the Relative Deprivation Scale.

• (4) Finally, participants were informed that they would conduct a market survey on consumption concepts. They needed to choose the consumption tendency of the product on the paper questionnaire. At the end of the whole study, the experimenters explained the real purpose of our study and gave participants a small gift in return for cooperation.

Statistical analysis

We adopted SPSS 26.0 for all statistical analyses in this study. The PROCESS macro developed by Hayes [54] in SPSS was applied to test the mediation effect. If the macro generates bootstrap confidence intervals that do not contain zero, it means reaching a level of significance.

Results

Manipulation check of passive social networking site use

Using an independent sample t-test, we found a significant difference in self-reported PSU between the PSU group (M = 4.63, SD = 0.69) and the non-PSU group (M = 1.43, SD = 0.46), t(201) = 38.88, p < 0.001, indicating an effective manipulation.

Preliminary analysis

Descriptive statistics (means and standard deviations) and correlation coefficients of each variable of Study 2 are presented in Table 2. According to the results, age was significantly positively associated with daily PSU (r = 0.147^*, p = 0.037). Manipulated PSU was positively related to RD (r = 0.386^**, p < 0.001) and CC (r = 0.259^**, p < 0.001). Daily PSU was positively correlated with CC as well (r = 0.200^**, p = 0.004). Moreover, RD and CC were positively correlated (r = 0.336^**, p < 0.001).

Table 2 Descriptive statistics and correlations for variables of study 2 (N = 203)

Mediation analysis

We used the PROCESS macro (Model 4) to examine the mediating role of RD between manipulated PSU and CC. The mediating effect was assessed by a bootstrapping procedure (5000 samples) that provided 95% bootstrapped confidence intervals (95% CI). The fundamental relationship between manipulated PSU and CC was significantly positive (β = 0.52, t = 3.80, p < 0.001). Further, the association between manipulated PSU and RD (β = 0.77, t = 5.94, p < 0.001) and the association between RD and CC (β = 0.28, t = 3.89, p < 0.001) were also positive and significant. When RD was included in the model, the relationship between manipulated PSU and CC remained significantly positive (β = 0.30, t = 2.13, p = 0.035). In addition, the total effect (effect = 1.007, 95% CI [0.485, 1.530]) was higher than the direct effect (effect = 0.591, 95% CI [0.043, 1.138]). RD partially mediated the positive correlation between PSU and CC (see Fig. 1).

Fig. 1

Study 2: Mediation model including passive social networking site use (predictor), relative deprivation (mediator), and conspicuous consumption (outcome)

Study 3

Method

Participants

In this part of the study, we recruited a total of 163 college students from Northwest China. There were 54 male students and 109 female students. The age range was 18 to 23 years old (M = 21.01, SD = 1.27). All participants provided informed consent.

Measures

Daily passive social networking site use

We measured this variable with the four items that we used in Study 1 and Study 2, and the Cronbach’s α of this scale in this part was 0.77.

Manipulation of passive social networking site use

Manipulation and manipulation check were the same as in Study 2.

Objective socioeconomic status

Our approach to measuring objective SES was similar to Kraus and Keltner [55]. SES was represented by the addition of standardized scores of paternal education, maternal education, and monthly family income. All our questions are in Chinese. A higher score indicates a higher objective SES.

Parental education was coded into five levels, 1 = never went to school, 2 = primary school graduate, 3 = junior high school graduate, 4 = senior high school graduate, 5 = junior college graduate/bachelor degree and above.

The monthly family income was divided into seven levels, 1 = less than 2,000 yuan, 2 = 2,001 ~ 3,000 yuan, 3 = 3,001 ~ 4,000 yuan, 4 = 4,001 ~ 6,000 yuan, 5 = 6,001 ~ 8,000 yuan, 6 = 8,001 ~ 10,000 yuan, and 7 = more than 10,000 yuan.

Manipulation of subjective socioeconomic status

We adopted the priming paradigm consistent with Kraus and colleagues [56]. Participants were first presented with a priming picture. The priming group with higher SSS viewed a picture of the bottom socioeconomic status (see Fig. 2), while the priming group with lower SSS viewed a picture of the top socioeconomic status (see Fig. 3). To improve the processing depth of the picture and increase the priming effects, the participants were asked to briefly describe how people live under these circumstances and the differences between themselves and these people after observation. After that, they were presented with a 10-step ladder (the MacArthur Scale) [55] (see Fig. 4). The steps of the ladder represent different social hierarchies, with 1 representing the bottom, where people live the worst off, have the lowest education, the least money, and the least respected jobs; and 10 representing the top, where people live the best off, have the highest education, the most money, and the most respected jobs. Participants were required to estimate the people in the pictures and their own socioeconomic status respectively, and the latter was considered as an indicator of manipulation check.

Fig. 2

Stimuli picture of priming group with higher SSS

Fig. 3

Stimuli picture of priming group with lower SSS

Fig. 4

The MacArthur Scale of subjective socioeconomic status

Relative deprivation

We measured this variable with the four items that we used in Study 2, and the Cronbach’s α of this questionnaire was 0.73 in this study.

Conspicuous consumption

We measured this variable with the two items that we used in Study 2, and the Cronbach’s α of these two items was 0.92 in this study.

Design and procedure

In this study, a 2 (PSU vs. non-PSU) × 2 (higher SSS vs. lower SSS) inter-group design was adopted, and participants were randomly assigned to 4 groups. PSU and SSS were both manipulated variables.

The PSU & higher SSS group contained 39 participants (14 males and 25 females); the non-PSU & higher SSS group contained 43 participants (15 males and 28 females); the PSU & lower SSS group contained 42 participants (13 males and 29 females); and the non-PSU & lower SSS group contained 39 participants (12 males and 27 females).

The specific process was as follows:

• (1) Participants filled in demographic information in the questionnaire and completed the Objective Socioeconomic Status Scale and the Surveillance Use Scale.

• (2) Participants’ SSS was manipulated. After observing the materials, they were required to assess the socioeconomic status of the people who live in the condition of the picture and their own by the MacArthur Scale.

• (3) PSU was manipulated. And participants also needed to answer related questions afterward to ensure serious engagement.

• (4) After then, participants completed the Relative Deprivation Scale.

• (5) Finally, participants conducted a market survey on consumption concepts. The experimenters explained the purpose of the study to participants and gave them a small gift at the end.

Statistical analysis

The statistical analysis was the same as in Study 2. In addition, we also tested moderated mediation effects by PROCESS macro in SPSS.

Results

Manipulation check of passive social networking site use

Also using an independent sample t-test, the PSU group (M = 5.28, SD = 0.70) and non-PSU group (M = 1.53, SD = 0.49) show a significant difference in the reported browsing of SNS information, t(161) = 39.57, p < 0.001, which proved that our manipulation of PSU was effective.

Manipulation check of subjective socioeconomic status

An independent sample t-test was used to examine the effectiveness of the manipulation of SSS. The results show that there was a significant difference in SSS score between the priming group with higher SSS (M = 5.17, SD = 0.77) and the priming group with lower SSS (M = 3.57, SD = 0.85), t(161) = 12.61, p < 0.001, indicating that the manipulation of SSS was effective.

Preliminary analysis

Table 3 lists the results of descriptive statistics and correlation analysis. It can be seen that neither gender nor age was significantly associated with other variables (p > 0.05). Manipulated PSU was positively related to RD (r = 0.195^*, p = 0.013) and CC (r = 0.278^**, p < 0.001). And daily PSU was positively correlated with CC as well (r = 0.158^*, p = 0.044). In addition, objective SSS had no statistically significant relationship with other variables, while manipulated SSS was negatively associated with RD (r=-0.179^*, p = 0.023) and CC (r=-0.234^**, p = 0.003). There was a positive correlation between RD and CC (r = 0.421^**, p < 0.001).

Table 3 Descriptive statistics and correlations for variables of study 3 (N = 163)

Mediation analysis

To test the mediating role of RD between manipulated PSU and CC, we used the PROCESS macro (Model 4) to assess the mediating effect by a bootstrapping procedure (5000 samples) that provided 95% CI. The results suggest that the fundamental relationship between manipulated PSU and CC was significantly positive (β = 0.56, t = 3.68, p < 0.001). Further, the association between manipulated PSU and RD (β = 0.39, t = 2.53, p = 0.013) and the association between RD and CC (β = 0.38, t = 5.35, p < 0.001) were also significantly positive. When RD was included in the model, the relationship between manipulated PSU and CC remained significantly positive (β = 0.41, t = 2.86, p = 0.005). In addition, the total effect (effect = 0.846, 95% CI [0.392, 1.301]) was higher than the direct effect (effect = 0.620, 95% CI [0.192, 1.049]). Consistent with the results of Study 2, RD partially mediated the positive correlation between PSU and CC (see Fig. 5).

Fig. 5

Study 3: Mediation model including passive social networking sites use (predictor), relative deprivation (mediator), and conspicuous consumption (outcome)

Moderated mediation analysis

The subsequent goal of this research was to investigate the potential moderating role of SSS in the direct and indirect (via RD) relationship between manipulated PSU and CC. A moderated mediation analysis was completed using PROCESS macros (Model 8) to obtain a bootstrapping procedure (5000 samples) with 95% CI for mediation models with different levels of SSS. The results are presented in Table 4.

Table 4 Results of moderated mediation analysis of study 3

In Model A, we estimated the moderating role of SSS in the relationship between manipulated PSU and RD. Manipulated PSU significantly and positively predicted RD (B = 0.44, SE = 0.12, p < 0.001). In addition, there was a negative interaction between manipulated PSU and SSS in predicting RD (B=-0.44, SE = 0.17, p = 0.011). Model B displays the moderating role of SSS in the relationship between manipulated PSU and CC. Both manipulated PSU (B = 1.10, SE = 0.30, p < 0.001) and RD (B = 0.85, SE = 0.19, p < 0.001) showed a significant positive predictive effect on CC. The interaction between manipulated PSU and SSS negatively predicted CC (B=-0.93, SE = 0.42, p = 0.030). Moreover, the direct effect and mediating effect of different levels of SSS are shown in Table 5.

Table 5 Direct effect and indirect effect under different levels of SSS of study 3

To more intuitively illustrate the moderating effects of SSS, we also presented simple slope plots (Figs. 6 and 7), which show that manipulated PSU had a significant positive predictive effect on RD (β = 0.36, p = 0.001) and CC (β = 0.45, p < 0.001) in the lower SSS prime group. However, in the prime group with higher SSS, neither the predictive effect of manipulated PSU on RD (β=-0.01, p > 0.05) nor on CC (β = 0.07, p > 0.05) reached a significant level. Overall, the results supported hypothesis three, that is, SSS moderated the relationship between PSU and RD and the relationship between PSU and CC.

Fig. 6

A simple slope plot of the moderating effect of SSS (subjective socioeconomic status) between passive social networking site use and relative deprivation

Fig. 7

A simple slope plot of the moderating effect of SSS (subjective socioeconomic status) between passive social networking site use and conspicuous consumption

Discussion

In today’s society, people’s consumption patterns and material values are changing dynamically with the development of the Internet. Specifically, there seems to be a strong link between the popularity of SNSs and the growth of CC. The present research focused on the relationship between PSU and CC among young people and further examined the mediating (RD) and moderating (SSS) effects between them.

Passive social networking site use and conspicuous consumption

Firstly, we found that PSU can positively predict CC, which is consistent with previous research that PSU is associated with materialism and irrational consumption behaviors [19, 27]. Our contribution was to take the experimental manipulation approach rather than the self-reported method, the former is more powerful to test causality.

Although SNSs essentially provide a platform for bilateral communication, we deem that it does not completely differ from traditional mass media (such as TV) since users can choose to just observe others and not display themselves. For this reason, the cultivation theory [24] also applies to SNSs. In other words, SNSs shape people’s material values and consumption behaviors (such as CC). When browsing on SNSs, users are more exposed to embellished information. Some of the content contains CC-related symbols [12, 13], which increases the accessibility of CC. Therefore, passive users are more likely to have an incentive to imitate upward, which promotes CC.

Furthermore, without controlling for specific content that was required to be browsing, we found that what is required to produce this effect is browsing the typical posts of participants. This may confirm the ubiquity of CC-related content on SNSs, allowing participants to be inadvertently exposed to CC-related content in their daily browsing. In a manner of speaking, SNSs, with their public, efficient, and timely feedback characteristics, have become important channels for people to display CC [12].

The mediating role of relative deprivation

Besides its behavioral effects (such as triggering unfavorable consumer behavior), PSU is also linked to a series of less healthy emotional and cognitive outcomes including envy and lower subjective well-being [14, 31, 32]. Our research suggests that PSU was closely related to RD as well, and the relationship between PSU and CC was mediated by RD.

Since the users of SNSs are committed to posting information that presents a good image, SNSs become a very concentrated representation of the enviable aspects of society. As a result, there is an opportunity to cause an automatic upward comparison when browsing SNSs [15, 23, 27, 30, 31], so as to experience the sense of RD [34]. As previous research has shown, frequent Facebook users tend to view other people as happier than they are and are more inclined to consider society as unfair [33]. People have various responses at this time, among which consumption behaviors become a means of compensating the psychological assets [28]. To a certain extent, the public tends to associate the possession of conspicuous goods with higher status, that is to say, owning such products gives individuals who are psychologically damaged vicarious satisfaction. For example, individuals under self-threat seek status goods to alleviate psychological distress [38]. Therefore, based on the motivation to reduce these aversive states (e.g., RD) during the passive use of SNSs, people seek relief from CC.

However, it must be admitted that our research included only young college students. In general, younger people are more sensitive than older adults to social behaviors that affect mood, such as social comparison [57]. Based on the theoretical perspective that the psychological effects of PSU are realized through social comparison, it can be assumed that the negative mental health impacts of PSU (e.g., RD) may be more pronounced in younger people than in older people. In fact, research has found that the positive association between PSU and mental ill-being (including anxiety, stress, and depression) becomes weaker as users age [58]. Therefore, although the psychological mechanism may be invariant, the effect size of the mediating role of RD may not be generalizable to all ages.

The moderating role of subjective socioeconomic status

In the final part of the research, we also investigated the moderating role of SSS under the framework of the differential susceptibility to media effects model [39]. The results show that the moderating effects of SSS were present both directly and indirectly (via RD) between PSU and CC. Specifically, higher SSS mitigated the feeling of RD and CC tendency in the PSU group compared with the non-PSU group. This is consistent with previous results that individuals with lower SSS are more apt to engage in CC [11, 42].

On the one hand, the perception of resource scarcity guides the individual’s attention preference [43, 44]. Individuals with lower SSS are more attracted to wealth-and-status-related information based on their lack of these resources. Consequently, conspicuous goods with symbolic meaning displayed on SNSs are more attractive to individuals with lower SSS, and then guide them to a higher tendency of CC.

On the other hand, lower SSS can also be a risk factor for RD and indirectly promote CC by enhancing RD. RD emphasizes resentful feelings and beliefs of unfairness arising from social comparisons [45, 59]. Being aware of being inferior to others leads to a feeling of being at a disadvantage, which results in negative emotional responses [60]. For lower SSS individuals, they are concerned about content with status signals on SNSs, which are often glorified. As a result, they are more likely to compare upward, and the strong contrast prompts them to feel status anxiety, which creates the sense of RD.

Research implications and limitations

Firstly, in the past few years, the influence of SNSs on people’s cognition, emotions, and behaviors has become a hot topic of research. However, the research on the specific use pattern of SNSs on users’ material values and consumption behaviors is insufficient. Our main finding is that PSU can promote CC. This result applies the cultivation theory of mass media to SNSs and expands our understanding of the effects of SNSs on users.

Secondly, PSU may influence CC tendency by inducing RD. This finding is mainly based on the compensatory consumption behavior model, in which people’s psychological needs may be impaired by exposure to positive information, prompting them to use products as a substitute for gratification. However, it is notable that this kind of compensatory behavior emphasizes symbolism, which does not directly address the source of self-loathing caused by a perceived unfavorable situation [28]. Therefore, this inspires us to beware of falling into the consumption trap in various social comparison situations, and instead compensate for psychological needs in healthier ways.

Finally, in terms of differential susceptibility to media effects, we found that higher SSS may be a protective factor for the adverse effects of SNSs on users. SSS does not rely entirely on objective indicators of SES, but rather on beliefs relative to others. This suggests that guiding people to think about what they already have, or what they have more than others may be an effective way to counteract the negative psychological effects caused by PSU.

In summary, our research contributes to the development of psychological intervention programs aimed at reducing people’s CC in the context of PSU (for example, interventions based on reducing the feeling of RD or raising SSS of users).

Meanwhile, there is still room for further exploration. Although this research explored the effect of PSU on CC, we did not record the specific content of users when browsing SNSs. Although we can assume that the typical browsing content contains CC-related content and is sufficient to promote CC, it can be further rigorously verified. For example, adding a survey of the types of content that participants browse as a supplement would be beneficial. In addition, considering that browsing different types of information (e.g., wealth, achievements, and physical fitness) may have different psychological effects on passive users, it would be interesting to investigate whether these kinds of information can promote CC and its internal mechanisms.

Moreover, since we conducted only one experiment on the moderating effect, it is necessary to re-examine the effects of SSS on various negative outcomes of PSU to consolidate our conclusion about individual differences in media susceptibility.

Finally, the participants in this research were college students. Even though they are a large group to use SNSs and are susceptible to Internet information, the representation of the participants was still insufficient. Although direct and mediated pathways may hold across age groups and cultural contexts, their effect sizes may not be consistent. For example, as we mentioned above, the effect size of the mediating role of RD may not generalize to all ages. Besides, our sample size was not large enough. Therefore, another shortcoming of our research is its limited external validity. This may mean that the relationship between PSU and CC needs to be investigated at other age stages and in different social groups before we can generalize our conclusions.

Conclusion

Overall, our research supported three hypotheses. Firstly, passive social networking sites use positively predicted conspicuous consumption. Secondly, relative deprivation played a mediator role between passive social networking site use and conspicuous consumption. Finally, subjective socioeconomic status directly and indirectly (via relative deprivation) moderated the relationship between passive social networking site use and conspicuous consumption.