Compulsive buying is a problematic behavior that could lead to negative outcomes (e.g., debts and negative relationships with relatives). Due to the COVID-19 pandemic, people changed their shopping behavior and experienced various life events. This research aims at exploring the relationship between life events and compulsive buying in Indonesia. A qualitative approach using in-depth interviews was employed. Nineteen participants who faced changing course of life, either positive (e.g., give birth) or negative (e.g., sudden death of family), during the pandemic were interviewed. We found that positive life events cause people to feel over-excited and over-grateful and shop compulsively. Negative life events that result in self-doubt and lower self-esteem also lead to a compulsive behavior. The study contributes to the compulsive behavior literature by exploring the antecedent of compulsive buying from the live events perspective and provide implications for policy makers and marketers.

Introduction Compulsive buying refers to a condition where consumers make purchases repetitively and excessively (Ridgway, Kukar-Kinney & Monroe, 2008; Japutra, Ekinci & Simkin, 2017). Previous literature shows that two types of behaviors characterize compulsive buying: impulsive buying and obsessive-compulsive buying (Ridgeway et al, 2008). Impulsive buying refers to unplanned purchase due to consumers’ inevitable impulse (Kacen & Lee, 2002), and obsessive-compulsive buying reflects a preoccupation in buying to reduce anxiety (Ridgway et al, 2008). Compulsive buying behaviors have been analyzed under the framework of motivation theory (McGuire, 1976). Nevertheless, research is still needed to understand more on the phenomena of compulsive buying (e.g., Kukkar-Kinney et al., 2016; Japutra et al., 2017). This study aims to explore the antecedents of compulsive buying behaviors using implicit theories. According to the implicit theories, mindset shapes the motivation of consumer behaviors (Dweck, 2000; Murphy & Dweck, 2016). Mindset refers to the beliefs about the nature of human characteristics, and individuals may possess two types of mindset – fixed and growth mindset (Murphy & Dweck, 2016). People with a fixed mindset believe that individuals’ qualities such as intelligence and competence are set and hard to change, whereas those with a growth mindset tend to believe that all individuals are able to change and develop through efforts and experiences. We argue that mindsets influence compulsive buying, and we propose that deal proneness mediates the relationship between mindset and compulsive buying. In doing so, we aim to enhance our knowledge in understanding how mindset affects compulsive buying behavior. Literature review and hypotheses According to the implicit theories, consumers with fixed mindsets believe that one’s abilities are fixed and hard to improve, and thus feel the need to prove, to themselves and others, that they have the abilities and/or they are successful (Murphy & Dweck, 2016). Thus, fixed mindset consumers tend to demonstrate their worth by using image-enhancing products and brands (Park & John, 2012). Deals, such as free gifts and offer of coupons, reduce the transaction cost and increase the perceived value of these image-enhancing purchases. In particular, since fixed mindset consumers favor success with little effort (Murphy and Dweck, 2016), deals can help them achieve their goal of image enhancement with lower cost. Thus, we hypothesize that: H1 Fixed mindset is positively related to deal proneness. For consumers with growth mindsets, a major motivation for their consumer behavior is to learn and improve (Murphy & Dweck, 2016). As the research by Blackwell et al. (2007) shows, growth mindset consumers are eager to participate in the self-improving process and achieve mastery. Thus, growth mindset consumers may perceive the information of discounts, free gifts and coupons as part of the adventurous process where they make the cost-benefit analysis and improve their abilities as wiser consumers. Thus, we propose: H2 Growth mindset is positively related to deal proneness. Previous studies show that compulsive buying is associated with high deal proneness (Kukar-Kinney et al, 2012). Deals may imply perceived value of the purchase and enhanced shopping enjoyment (Grewal, Monroe, & Krishnan, 1998), and thus serve as an effective contextual factor in inducing compulsive behaviors (Kukar-Kinney et al, 2016). Furthermore, deals provide an excuse and rationale for the purchase, which can be used to overcome the sense of guilt compulsive buyers often experience after their compulsive buying behavior (O’Guinn & Faber, 1989). Thus, we make the following hypotheses. H3 Deal proneness is positively related to impulsive buying. H4 Deal proneness is positively related to obsessive-compulsive buying. Finally, we argue that deal proneness mediates the relationship between consumer mindsets and compulsive buying behavior. According to the implicit theories, consumer mindsets inspire how consumption goals are pursued (Murphy & Dweck, 2016). Consumers with a fixed mindset pursue a performance goal, and they tend to use brands to feel positive about themselves and improve impression on others (Park & John, 2010). In contrast, consumers with a growth mindset hold that people can always learn and improve and thus are tuned to learning goals (Murphy & Dweck, 2016). Hence, for fixed mindset consumers, deal offers suggest lower costs for image-enhancing purchases, and provide an excuse for the compulsive buying behavior. For growth mindset consumers, deal offers can imply a learning and adventurous process .These consumers may feel that they can make better purchasing decisions by taking advantage of various deals. We thus hypothesize that: H5 Deal proneness mediates the relationship between fixed mindset and impulsive buying (H5a), between fixed mindset and obsessive-compulsive buying (H5b), between growth mindset and impulsive buying (H5c), and between growth mindset and obsessive-compulsive buying (H5d). Method A questionnaire was developed to gather responses and test the hypotheses. All of the items to measure the constructs were developed from existing scales based on previous research. Fixed and growth mindsets were measured using scales developed by Park and John (2012). Deal proneness was measured using items following Lichtenstein et al. (1997). Impulsive buying and obsessive-compulsive buying were measured using items developed by Ridgway et al. (2008). All items were rated on a 7-point Likert scale, ranging from “strongly disagree” (1) to “strongly agree” (7). The questionnaire was administered using an online survey (N=421 respondents). Of these, 57.7% were female, 71.5% had a university degree, 50% were 31-40 years old, 41% were 26-30 years old, and 46.3% had a monthly income of 5,001-10,000 RMB. Results and discussion To test the hypotheses within the research model, a Structural Equation Modeling (SEM) approach was employed, using AMOS 18.0. First, a measurement model was created to assess the validity and reliability of the scales. The distribution of the data was checked. The absolute value of the skewness and kurtosis of each items were within +/- 1, suggesting normal distribution was achieved. The measurement model produced good fit (Hair et al., 2010): χ2(109) = 281.21, χ2/df = 2.58, GFI = .93, NFI = .93, CFI = .96, and RMSEA = .06. All values representing the AVE were greater than 0.5 and greater than the squared inter-constructs correlations, indicating convergent and discriminant validity were achieved (Fornell & Larcker, 1981). Cronbach’s alpha values exceeded .70, indicating the constructs were reliable (Hair et al., 1995). The results of the checking common-method variance problem through exploratory factor analysis (EFA) test revealed 3 factors with Eigen values greater than 1. The results accounted for 64.67% of the total variance, where the first factor accounted for 27.55% of the total variance, suggesting that common-method variance did not pose a significant problem since there was no general factor in the un-rotated structure (Podsakoff et al., 2003). Next, a structural model was created. The structural model produced good fit (Hair et al., 2010): χ2(114) = 476.15, χ2/df = 4.18, GFI = .89, NFI = .89, CFI = .91, and RMSEA = .09. Table 1 displays the results of SEM. The results support H1 and H2. Both fixed and growth mindsets are positively associated with deal proneness. The results support H3 and H4, which shows that deal proneness are positively associated with impulsive and obsessive-compulsive buying. The results support H5a, which states that deal proneness mediates the relationship between fixed mindset and impulsive buying. However, the results do not support H5b, H5c and H5d. Conclusion Using the implicit theories, this research aims to gain better insight into compulsive buying behavior. Our findings, obtained from a sample of respondents in China, show that deal proneness serves as a mediator between fixed mindset and compulsive buying behaviors. According to the implicit theories, consumer mindsets inspire how consumption goals are pursued (Murphy & Dweck, 2016). For instance, consumers with a fixed mindset pursue a performance goal. They tend to use brands to feel positive about themselves and improve impression on others (Park & John, 2010). Thus, it is likely that fixed mindset consumers buy compulsively to signal and communicate their “self” to others. In particular, for fixed mindset consumers, deals may increase the perceived value of image-enhancing purchases. Thus, deals provide an excuse for the compulsive purchase where fixed mindset consumers can improve self-image and demonstrate their worth with lower costs. Given this mediating role of deal proneness between fixed mindset and compulsive buying, it will be interesting to test further how consumers with a fixed mindset respond to different types of deals in future research. For firm managers and public policy makers, our findings imply that, to lessen consumers’ overspending, firms should reduce excessive number of deals, and governments should also regulate firms’ advertisement so that it will not overly promote deals.

Many studies in the marketing literature focus on investigating the effects of brand attachment and self-congruence on favorable consumer behavior, such as brand loyalty, positive Word-of-Mouth, resilience to negative information (Thomson et al., 2005, Japutra et al., 2014a). However, limited studies examine the relationship between brand attachment, self-congruence and negative consumer behavior. For instance, Johnson et al. (2011) posit that when consumers are in relationships with brands and more self-relevant, they are more likely to induce retaliatory actions when the relationship ends. Highlighting the dark side of brand attachment, Japutra et al. (2014b) show that stronger brand attachment leads to a wide range of unwanted consumer behaviors including trash-talk, Schadenfreude, and anti-brand actions. These studies show how brand attachment is capable of hurting consumers. However, only limited studies pay attention on the negative consequences of brand attachment (e.g. Johnson et al., 2011; Japutra et al., 2014b). The purpose of this study is to investigate the effect of self-congruence and brand attachment on compulsive buying behavior.

The concept of retail therapy which refers to “the phenomenon in which consumers buy things to make themselves feel better” (Kacen, 1998) has been used by the U.S. retail firms in marketing (Kang & Johnson, 2010). Previous studies found that retail therapy concept was associated to consumption behaviors including compulsive buying (Faber & O’Guinn, 1992; O’Guinn & Faber, 1989). Frost et al. (1998) found that compulsive hoarders show compulsive buying tendencies. O’Guinn and Faber (1989) found that compulsive buyers are more likely to confirm compulsivity as a personality trait with a low self-esteem. Lack of research into association of retail therapy to compulsive consumer behaviors motivated us to pursue this study which investigate relationships among retail therapy, compulsive buying and compulsive hoarding, and to examine the moderating effect of consumers’ personality traits on the relationship between retail therapy and compulsive hoarding. Our research questions are: (1) How retail therapy is related to compulsive buying and compulsive hoarding? and (2) Which personality traits moderate the relationship between retail therapy and compulsive hoarding? Based on the literature review, the following hypotheses were proposed. H1: Retail therapy is positively related to compulsive buying; H2: Compulsive buying is positively related to compulsive hoarding; H3: Retail therapy is positively related to compulsive hoarding & H4: Personality trait moderates the relationship between retail therapy and compulsive hoarding. The researchers used 12 items to measure retail therapy (Kang, 2009), 6 items to measure compulsive buying (Faber & O’Guinn, 1992), 9 items to measure compulsive hoarding (Frost, Sketekee, & Grisham, 2004) 8 items of Big Five Inventory (Rammstedt & John, 2007) and 15 item Narcissistic Personality Inventory (NPI) (Ames, Rose & Anderson, 2006) to measure personality trait based on 7‑point Likert scales (1 = strongly disagree, 7 = strongly agree). The online survey was conducted with college students enrolled at a major Midwestern University. A total of 354 undergraduate and graduate students’ responses were used to analyze data. A principal component factor analysis with varimax rotation was conducted to evaluate the measurement of each construct. The factor loadings were above of 0.60, indicating acceptable convergent validity. Reliabilities with Cronbach’s alphas for retail therapy,compulsive buying, compulsive hoarding and personality trait construct were 0.96, 0.81, 0.90 and 0.81 respectively, exceeding the suggested level of 0.70. Simple linear regression was performed to test hypotheses. The results revealed that F statistic (F(1, 352)=126.53, p =.000) was significant and regression coefficient was statistically significant. Therefore, H1 was supported and consistent with previous findings (Faber & O’Guinn, 1992). The regression coefficient for the path from compulsive buying to compulsive hoarding was statistically significant (F(1, 352)=77.15, p=.000). Therefore, H2 was supported and consistent with Frost et al. (1998)’s study. The retail therapy contributed significantly to the prediction of compulsive hoarding with the statistically significant regression results (F(1, 352)=20.28, p=.000), supporting H3. From the principle component analysis with varimax rotation, four factors were extracted from 23 personality trait measurement items which are labeled as “authority-leadership narcissism”, “self-esteem narcissism”, “positive disposition”, and “negative disposition”. All factor loadings were above 0.60 for their respective factors. Then, the researchers divided one single group into four personality trait groups, using these four factors. 4%, 27.1%, 53.7% & 15.3% participants represented “authority-leadership narcissism”, “self-esteem narcissism”, “positive disposition” & “negative disposition” group respectively. Simple regression analysis was performed to test H4. The regression coefficients for positive disposition group (F(1, 188)=13.19, p=.000) and negative disposition group (F(1, 52)=5.01, p≤.05) were statistically significant. H4 was partially supported. The results indicated that people from two groups characterized by positive/negative personality tend to engage in compulsive hoarding to alleviate their negative feelings or mood. Anyone with narcissistic personality with a high self-efficacy may not be engaged in compulsive hoarding even when a therapeutic treatment is made through shopping. They tend to purchase products for changing their moods, but their therapeutic behavior doesn’t lead to compulsive hoarding, which causes sufferings in the end. The low self-esteem has been presented with people being engaged in compulsive behaviors (Marlatt et al., 1988). These findings can help apparel marketers develop the strategies to upgrade their sellingenvironment entertaining so that their customers’ moods are repaired and customers feel satisfied through therapeutic shopping behavior. This study has a limitation that prevents us from generalizing the results to the young consumer population due to sample size to college students.

Research findings show that money attitude dimensions variedly affect compulsive buying. We surveyed 750 Generation Y South Africans to examine whether gender and family resources received during childhood is influencing the varied impact of money attitude dimensions on compulsive buying. Depending on whether low or high family resources were received during childhood, we found gender differences and a similarity on how money attitude dimensions affect compulsive buying. In terms of similarity, we found that high provision of both tangible and intangible family resources during childhood promotes the development of budget money attitude, which negatively affects the development of compulsive buying behavior.