POLITICAL IDEOLOGY AND CUSTOMER FEEDBACK: DO CONSERVATIVES PROVIDE MORE VALUABLE FEEDBACK TO FIRMS? By Xiaoxu Wu A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Business Administration – Marketing – Doctor of Philosophy 2024 ABSTRACT Companies are eager to gather and utilize customer feedback about their core offerings, as it can provide a deeper understanding of customer sentiment and serve as a source of competitive advantage. However, relatively little is known about the factors that drive customers to share managerially valuable feedback with the companies from which they purchase. Analyzing both large-sample customer survey data and experimental data across distinct consumer domains, we demonstrate that customers’ political ideology is an important factor influencing the value of their feedback to firms. Specifically, we find that the more conservative the customer, the more valuable their feedback is to companies. This is because customers who hold more conservative ideologies tend to have higher trust in the private sector, making it easier for companies to build trust with these customers. We also show that the effect of customer political ideology on the value of customer feedback is attenuated for large firms and strengthened for firms that receive external recognition (e.g., third-party awards). These findings contribute to existing research on both the marketing-political identity interface and customer feedback, offering implications for marketing managers and guiding future research and theory. Copyright by XIAOXU WU 2024 ACKNOWLEDGEMENTS To my mother and father: Thank you for believing in me and supporting me to pursue my dreams. The strength I've drawn from your love has been the guiding lights on my path. To Ben: Thank you for being there for me through every moment of excitement and every challenge. You are my constant source of motivation and happiness. To Dr. Basuroy: Thank you for always encouraging me to explore, learn, and improve. Your guidance has opened new horizons for me. To Dr. Pansari, Dr. Morgeson, and Dr. Park: Your support and guidance have been invaluable on my journey to becoming a better scholar. I am deeply grateful for your mentorship and for always being there for me. To Udit and Michael: I couldn't have asked for better friends to navigate the ups and downs of this PhD adventure. Thank you for every moment of laughter, support, and encouragement. iv TABLE OF CONTENTS INTRODUCTION ...........................................................................................................................1 RESEARCH BACKGROUND AND THEORETICAL FRAMEWORK ......................................7 METHODS AND RESULTS ........................................................................................................19 DISCUSSION AND FUTURE RESEARCH ................................................................................37 REFERENCES ..............................................................................................................................45 APPENDIX A: CUSTOMER FEEDBACK VALUE CODING SCHEME .................................57 APPENDIX B: CUSTOMER FEEDBACK VALUE PREDICTION ..........................................60 APPENDIX C: FIGURES ............................................................................................................62 APPENDIX D: TABLES ...............................................................................................................71 v INTRODUCTION Encouraging customers to provide feedback is an important goal for marketing professionals, as firms can utilize this feedback to improve their products and services, adapt to marketplace dynamics, and remain competitive (Pansari and Kumar 2017). As Microsoft founder Bill Gates famously suggested, “We all need people who will give us feedback. That’s how we improve.” Customer feedback is valuable for companies not only because it enables them to anticipate and adapt to marketplace dynamics, but also because it is a vital conduit of information for the product, service, and/or business model “open innovation” currently pursued by industry-leading global firms like Apple, Microsoft, Starbucks, Procter & Gamble, and innumerable others (e.g., Chesbrough and Appleyard 2007; Randhawa, Wilden, and Hohberger 2016). Furthermore, in an age marked by increasing consumer power, customers now demand opportunities to offer their insights to firms, both positive and negative. Indeed, 78% of U.S. consumers indicate a preference for brands that collect and accept feedback (Microsoft 2017). Firms are increasingly utilizing marketing research as a tool for knowledge development and are investing considerable resources in gathering customer feedback. This includes investments in building in-house research capabilities to collect customer feedback through surveys and outsourcing such work to external market research vendors. In 2021 alone, global revenue for the market research industry exceeded $76.4 billion (Statista 2022). However, the common issues that many firms and market research agencies encounter include low response rates and low-quality customer feedback (Forbes 2022; Ode 2016). Therefore, identifying customers who not only provide feedback but also provide high-quality and valuable feedback is critically important. Surprisingly, little is known about the factors that drive customers to provide truly valuable feedback, i.e., feedback that is useful for firms in terms of improving their 1 products and services. Providing feedback to a firm requires an investment of customers' time and energy, and as such, it reflects a customer’s commitment to the relationship with the firm. While customers can provide positive or negative feedback, in both instances, feedback can be viewed as a relationship-preservation strategy (Umashankar, Ward, and Dahl 2017). That is, providing positive feedback represents a customer’s reward to the firm and an encouragement to continue delivering satisfying products or services in the future (Chen et al. 2023), while providing negative feedback (e.g., suggestions for improvement or failure-related complaints) represents the customer’s attempt to repair a threatened relationship so that it might endure (Umashankar, Ward, and Dahl 2017). Hence, factors that motivate customers to form strong relationships with firms might also drive customer feedback behaviors. Recently, one newly investigated customer- level factor and its relationship with myriad customer attitudes and behaviors have drawn increased interest in the marketing literature: political ideology, defined as a “set of beliefs about the proper order of society and how it can be achieved” (Erikson and Tedin 2015; Jost 2017). Though political ideology has been shown to shape both customer preferences (Khan, Mishra, and Singh 2013; Paharia and Swaminathan 2019) and post-consumption evaluations (Fernandes et al. 2022; Jung et al. 2017), very little is known about how customers’ political ideology impacts their participation in a firm’s knowledge development process through feedback. Considering that political ideology fundamentally reflects and reinforces individuals’ relational motives (along with epistemic and existential motives; Jost, Nosek, and Gosling 2008; Jost, Federico, and Napier 2009),1 in the current study, we seek to answer the following question: 1The epistemic motives concern the need for cognition, evaluation, and cognitive closure, and political ideology serves such motives by offering certainty. The existential motives concern denial of death/anxiety, threat management, and coping with emotional disgust, and political ideology serves such motives by offering security. 2 How do consumers’ political ideologies affect the value of the feedback that they provide to firms? To examine this question, we build a conceptual framework that draws on and expands the extant research on political ideology and customer feedback in marketing, and we empirically test this relationship across three studies. In Study 1, we examine a unique and large dataset of 15,953 current customers of 117 firms across five service-dominant industries over the 2016- 2020 period. Accounting for a variety of relevant factors, addressing endogeneity through an instrumental variable strategy, and using multiple measure and estimation approaches, we find that customers who are more conservative (vs. liberal) provide more valuable feedback to firms. This effect is attenuated for larger firms and strengthened for firms that receive external recognition for their products and services. In Studies 2a and 2b, we show that the effect of customer political ideology on customer feedback value occurs because conservative customers have higher trust in the private sector, which predisposes them to trust the firm and facilitates the building of higher trust among these customers. Through this research, we advance the customer feedback literature in two ways. First, we identify customer political ideology as a driver of customer feedback value and demonstrate that the heterogeneity in customer feedback value partially stems from customers’ characteristics as rooted in their political ideologies. Second, we expand upon existing findings from prior literature, which suggested that conservatives are less likely to complain (a form of feedback) about a firm to a regulatory body (Jung et al. 2017), by showing that conservatives are, in fact, more likely to provide valuable feedback (both positive and negative) directly to the firm because of underlying motives to help firms improve their offerings (versus the motive of The relational motives concern political socialization, social identification, and the need for shared reality, which an ideology can provide through solidarity. 3 externalizing anger via complaints to regulatory bodies). Third, to the best of our knowledge, prior research on customer engagement has only used self-reported measures of feedback value developed by Kumar and Pansari (2016). We complement their work by developing and validating a behavioral measure of customer feedback value that captures the usefulness of customer feedback in terms of helping managers improve firms’ products and services. Drawing on existing literature on knowledge management (Clarkson, Janiszewski, and Cinelli 2013; Prabhu, Chandy, and Ellis 2005) and interviews with industry experts, this new measure expands the conceptualization of customer feedback value by delineating its two dimensions: breadth and depth. Developing a measure that reliably captures the value of customer feedback is beneficial because it not only enables scholars to further study the antecedents and outcomes of customer feedback value but also allows marketers to identity the most valuable feedback to aid their product and service development and improvement efforts. Furthermore, this research makes two significant contributions to the burgeoning political identity literature in marketing. First, among the three classes of psychological factors that form the motivational sub-structure of an individual’s political ideology, research in marketing has predominantly invoked epistemic and existential motives to examine how consumer political ideology affects certain attitudes and behaviors (e.g., Fernandes et al. 2022; Kidwell, Farmer, and Hardesty 2013; Jung and Mittal 2021). However, prior research falls short of elucidating the potential implications of the relational motives that underlie consumers’ political identities. In response to this gap, we identify the value of customer feedback to firms as an outcome attributable to the relational motives underpinning their political identity. We demonstrate that the need for a shared reality, as manifest in conservatives’ (versus liberals’) shared belief in the 4 most efficient organization of society through an unfettered free market or the private sector (versus the government sector), drives them to place higher trust in firms (Pew 2019). Since providing valuable feedback entails a willingness of customers to cooperate with a firm, and since higher trust drives cooperation between exchange partners (Morgan and Hunt 1994), we show that higher trust in firms leads customers to provide more valuable feedback. Second, we deepen this understanding by identifying a context that challenges (firm size) and a context that reinforces (external firm recognition) the trust-based mechanisms underlying the effect of conservatism on feedback value as moderators. Prior research in marketing has looked at the objective liabilities of firm size from the firm perspective, such as challenges to the actualization of returns from innovation and the ability of firms to satisfy a heterogenous customer base (Marinova 2004; Bhattacharya, Morgan, and Rego 2022). We demonstrate a unique perception-based challenge posed by firm size from the customer perspective. While conservatives have higher trust in firms because of their pro-free market beliefs, such trust is diminished for larger firms, as large firms tend to be perceived as more monopolistic and thus as antithetical to free market exchanges (Pew 2021). Next, we show that external recognition received by firms (e.g., third-party customer service awards) acts to strengthen the positive relationship between conservatism and feedback value. This is because external awards and recognition are high-quality signals of the often-invisible intent of firms to improve their products and services, and reinforce conservatives’ higher (relative to liberals) baseline trust in firms (Gemser, Leenders, and Wijnberg 2008). To the best of our knowledge, no prior work in marketing has linked these factors (firm size and external firm recognition) to consumers’ political ideologies, perhaps due to the aforementioned inadequate examination of the relational- motives-based cognitive substructure of political ideology. 5 Finally, our research has important implications for how firms seek and manage customer feedback. Given that customer political ideology is more observable than most psychological factors and can be more readily inferred using public information such as voter registration, election polling data, and social media (Duggan and Smith 2016; Jung et al. 2017), or easily measured through customer surveying, it is relatively more practical and efficient for managers to identify and target customers based on their political ideology. Additionally, our research demonstrates that managers can utilize natural language processing (NLP) algorithms to manage and analyze large samples of customer feedback data, which many firms collect in their customer surveys, and channel their energies to feedback that contains the highest feedback value. The remainder of the paper is structured as follows: In the next section, we provide the background of this study and review the extant literature on customer feedback and political ideology. Following this, we outline the theoretical foundations of the research, focusing on how divergent beliefs about trust in the private sector among conservatives versus liberals affect the customer trust-building process and further impact customer feedback value. We further discuss the moderating role of firm size and external firm recognition in these relationships. We then present the three studies that empirically test these relationships. Finally, we provide a summary and discussion of our findings, along with the implications of our results for both academic research and marketing managers. 6 RESEARCH BACKGROUND AND THEORETICAL FRAMEWORK Marketing Literature on Customer Feedback We define customer feedback as customers’ complaints, compliments, or thoughts about a firm’s products and/or services that are voiced directly to the firm, posted to public platforms, or provided to other institutions such as regulatory bodies (Celuch, Robinson, Walsh 2015). This definition highlights two core characteristics of customer feedback. First, customer feedback can be positive, negative, or mixed. Though all feedback can be extremely valuable to firms, the vast majority of prior research on customer feedback has focused on negative feedback (Bone et al. 2017; Gelbrich and Roschk 2011). Second, customers can provide their feedback through different channels: direct feedback through customer surveys (Bone et al. 2017; Morgeson et al. 2020), to employees during service encounters (Hekman et al. 2010), by posting opinions on public platforms such as social media, by offering online reviews (Wu et al. 2015), or through official complaints to regulatory bodies (Jung et al. 2017). Importantly, the choice of feedback channels may itself signal very different customer motivations. For example, prior literature has identified four main purposes for customer complaints: to obtain restitution, to vent anger, to help improve the service, or for altruistic reasons (Lovelock et al., 2008). A customer who voices a complaint directly to the firm is more likely to seek restitution or to help the firm improve its product or service; similarly, a customer who files a complaint with regulators or posts negative feedback online is more likely to vent anger. In the current research, we focus only on the customers’ direct feedback to firms, as firms have the greatest control over this channel. The preponderance of extant literature on customer feedback focuses on the customer- level outcomes of feedback behavior (e.g., Bone et al. 2017; Chen et al. 2023; Umashankar, Ward, and Dahl 2017) and firm-level outcomes of soliciting and/or incorporating customer 7 feedback (e.g., Beckers, van Doorn, and Verhoef 2018; Challagalla, Venkatesh, and Kohli 2009). Surprisingly, very few studies have explored the drivers of customer feedback. In the service marketing context, employees’ customer-orientation behavior (Celuch, Robinson, Walsh 2015) and preferential treatment of customers (Lacey, Jaebeom, Morgan 2007) have been shown to positively influence customer feedback behavior. Usefully, Jung et al. (2017) explored the effects of political ideology on customer complaints and demonstrated that conservative customers are less likely than liberal consumers to report complaints or to dispute complaint resolution efforts due to stronger motivations to engage in system justification. However, their study focuses on customer complaints (negative feedback) filed with regulators (e.g., the Consumer Financial Protection Bureau), feedback behaviors most likely motivated by a need to vent anger, whereas direct customer feedback to firms (as discussed earlier) is motivated by restitution-seeking or intentions to help firms improve their offerings. Additionally, although customer feedback is important, not all feedback is equally valuable for firms, because not all feedback is “robust, detailed, and useful.” As such, and to the best of our knowledge, this is the first research that investigates customer-level factors that drive the value of customer feedback (both positive and negative) from a managerial perspective. Marketing Literature on the Role of Political Ideology in Customer Behavior Recently, one relatively new factor and its relationship with customer attitudes and behaviors have drawn increased attention within the marketing literature: political ideology. With rapidly growing political polarization in the United States and throughout the world, political ideology has become a more salient feature of individual identities over the last two decades (Jost 2006; Lardieri 2020). A growing number of studies in marketing examine customers’ political ideology as a potentially significant factor influencing their behaviors (e.g., 8 Fernandes et al. 2022; Hydock, Paharia, and Blair 2020; Ulver and Laurell 2020). In line with the predominant theme of political psychology over the last few decades, most of these studies look at the impact of political ideology – typically positioned in the U.S. context on a “liberal vs. conservative” spectrum – as driving differential consumer preferences (e.g., Fernandes et al. 2022; Jung and Mittal 2021). These extant studies are based on the premise that underlying and motivating a customer’s political ideology are psychological factors that differentiate these ideological groups (Jost et al. 2003; Jost 2017). As discussed earlier, there are three categories of such factors that comprise the motivational substructure of political ideology: epistemic motives, existential motives, and relational motives (Jost 2009). A plurality of these studies in marketing have examined outcomes that are driven by epistemic motives, which largely concern certainty- seeking. For example, conservatives (versus liberals) are observed to have a preference for luxury goods and established national brands because of their stability-seeking tendencies (Kim et al. 2018; Khan, Misra, and Singh 2013). Several studies have identified outcomes traceable to existential motives, which concern security-seeking. For instance, social dominance orientation, a trait linked to security-seeking through dominance (Schwartz & Boehnke 2004), drives conservatives to take larger financial risks (Choma et al. 2014; Han et al. 2019) and prefer vertically differentiated products (Ordabayeya and Fernandes 2018). However, to the best of our knowledge, only one study in marketing identifies an outcome driven by relational motives, which concern solidarity-seeking. Winterich, Zhang, and Mittal (2012) find that conservatives (versus liberals) tend to donate more to charities managed by private entities (versus the government) because of their preference for capitalism. We extend their work and propose that the need for a shared reality, as manifest in conservatives’ (versus 9 liberals’) shared belief in the most efficient organization of society by means of an unfettered free market or the private sector (versus the government sector), drives them to place higher trust in firms (Pew 2019). Since providing valuable feedback entails a willingness of customers to cooperate with the firm and since higher trust drives cooperation between exchange partners (Morgan and Hunt 1994), we propose that higher trust leads customers to provide more valuable feedback to firms. Political Ideology and Trust in the Private Sector Political ideology is defined by and influences a wide variety of individual-level beliefs and perceptions that collectively inform a normative view of society. Yet, holistically, political ideology is often understood vis-à-vis an individual’s attitudes toward the major social institutions – the economy, government, education, family, and religion (Solak et al. 2012). That is, individuals’ beliefs about these institutions drive their ideological categorizations (e.g., ideology can be and often is inferred from how individuals perceive these institutions (Everett 2013)), and in turn, one’s (perhaps evolving) ideology can drive and result in adjustments to beliefs about these institutions and how they ought to function to achieve a “proper order of society.” Categorizations and definitions of political ideology differ across countries and political systems and are deeply interwoven with the unique political history and culture of diverse nations (Nilsson and Jost 2020). In the United States, the two predominant ideological groupings are liberals and conservatives. One key differentiator between these groups is their differing perspectives on the economy. Traditionally, American liberalism has been associated with those who believe in a well-regulated market economy, marked by a stronger central government and robust government intervention in the economy (e.g., Alterman 2013; Reeves 2018). 10 Additionally, liberals tend to place the collective good above a strict perspective on individual liberty – including, as we outline below, in their perceptions of liberty vis-à-vis property rights and the private sector (Alterman 2013; Reeves 2018). American conservatism, on the other hand, is associated with those who believe in an unfettered free market economy and minimal government intervention in the economy (Farmer 2005; Howison 2018; Niskanen 1988). Of particular note, conservatives tend to favor a strong position on individual rights and liberties, free from external (and especially governmental) interference, and are pro-business and pro- capitalism (i.e., “economic individualism”). Rooted in their conflicting perspectives on individual liberty and the economy, liberals and conservatives differ on two significant aspects most relevant to this research. While American liberals conceptualize individual liberty as the freedom to pursue unique “life projects” and access to the resources needed to do so (i.e., individuality through “positive liberty”), conservatives mostly define liberty as the absence of external obstacles and interference (i.e., individualism through “negative liberty”) (Berlin 2017). Indeed, within the American conservative tradition, liberty as individualism emerges through the concept of natural rights – rights that require protection from external infringement and interference (Locke 2013). As the government seeks to regulate and interfere with these exchanges (through taxation, appropriation, and coercion), it thus infringes upon natural rights, property rights, and liberty, generally viewed negatively by conservatives. Further for conservatives, this results in robust opposition to government intervention (limited government, checks on government power, etc.), and, contrarily, a strongly positive view of the private sector and the economy as the arena where individuals truly exercise their liberty by engaging freely in voluntary property exchanges with 11 others (Macpherson 1951).2 The differentiation of liberalism and conservatism vis-à-vis the major social institutions – particularly in terms of perceptions of liberty, the economy, and the private sector – provides a foundation for understanding the differences between these predominant political ideological groupings with respect to their trust in the private sector. Broadly, trust is defined as an individual’s perception of, and belief in, the reliability, truth, and strength of someone (i.e., an individual) or something (i.e., an institution) (Devos, Spini, and Schwartz 2002). In the political context, trust is often proposed to exist (or not) vis-à-vis institutions, including both formal governmental entities (e.g., the presidency, Congress, etc.) and somewhat more informal institutions like the free market and the private sector (Levi and Stoker 2000). Given the above discussion regarding the connection between conservatism and more positive perceptions of the private sector as an arena of free economic exchanges that produces outcomes preferable to those of government, along with the related empirical data in this domain (see footnote 2), we anticipate that conservatives have greater trust in the private sector relative to liberals. Trust in Private Sector and Trust in the Company Drawing on prior research, we further argue that conservatives’ greater trust in the private sector will translate to higher initial trust towards the firm from which they choose to purchase. This initial trust is normally formed prior to the customer having any first-hand experience with 2Voluminous empirical evidence supports this assertion, including recent data from the Pew Research Center on the beliefs and attitudes of political ideology groupings (Pew 2021). When asked the question “do business corporations make too much profit,” more than 75% of liberal-leaning respondents agree, while only about 25% of conservative- leaning respondents do so. Similarly, around 80% of conservative-leaning respondents believe that “government is doing too many things better left to businesses and individuals,” whereas even larger percentages of liberal-leaning respondents believe that “government should do more to solve problems” and that “the economic system in this country unfairly favors powerful interests” (Pew 2021). The Pew study segments both conservatives and liberals into four sub-categories and because these sub-categories tend to be closely grouped in their responses, we generalize and estimate the overall results (percentages) provided here. Distinct data from the American National Election Study suggests similar dynamics, with ideological conservatives expressing substantially more positive feelings towards both “big business” and “capitalism” than liberals (ANES 2021). 12 the firm (Berg, Dickhaut, and McCabe 1995; Kramer 1994). This initial trust plays an important role in the trust-building process with the firm, especially at the beginning of relationship formation when there are more risks and uncertainties involved, and influential events may affect the continuance of the relationship (McKnight and Chervany 2006). For example, initial trust among customers is critical for the survival of digital innovations by start-ups (Konya-Baumbach et al. 2019). Therefore, we anticipate that conservatives' greater trust in the private sector will positively influence their trust in the firms both initially and throughout their relationship with the firms. Trust in the Firm and Feedback Value For the last link of this proposed causal chain, we draw on the commitment-trust (CT) paradigm to discuss the role of customers’ trust in the firm in driving the value of feedback they provide (Morgan and Hunt 1994). According to CT, effective cooperation in economic exchange, defined as situations in which parties work together to achieve common goals (Anderson and Narus 1990), is based on norms of trust (Berry and Parasuraman 1991; Morgan and Hunt 1994). Once trust is established, it drives an exchange partner’s cooperation with the firm directly and through commitment, defined as an exchange partner’s belief that an ongoing relationship with the other is critical enough to warrant maximum effort at maintaining it (Morgan and Hunt 1994; Moorman, Zaltman, and Deshpande 1992). This is because trust enables ongoing learning and adaptation within an exchange relationship, with lasting effects on relationship performance (Palmatier et al. 2013; Selnes and Sallis 2003). Providing feedback to firms is an individual customer’s effort to help the firm improve its products and services, something that entails an investment in a relationship (Kumar and Bhagwat 2010). As such, and viewed from the CT perspective, providing valuable feedback is an 13 act of cooperation (Lacey, Suh, Morgan 2007) driven by a customer’s desire to maintain or improve the relationship with the firm. Several strands of recent research also allude to the role of trust in driving feedback. For example, Pansari and Kumar (2017) propose that the higher the level of trust customers have, the more willing they may be to engage with the firm in multiple ways, including by providing feedback. Umashankar, Ward, and Dahl (2017) find that the presence of trust between customers and service providers enhances the willingness of customers to voice their feelings to the firm. Furthermore, since an exchange relationship is built on trust and commitment, we expect customers to be more motivated to invest time and energy to provide useful feedback to the firms. Therefore, it is reasonable to suggest that consumers’ trust in the companies drives the value of the feedback they provide as a function of their commitment to a desired ongoing relationship with them. Based on the totality of the above theorizing, we hypothesize that: H1: Ideologically conservative customers (versus liberal customers) provide higher-value feedback to firms. H2: The association between conservative (versus liberal) customers and higher (versus lower) customer feedback value is mediated serially through greater trust in both the private sector and in firms. Moderator: Firm Size As noted earlier, while political ideology and the differences between liberals and conservatives provide useful categorization and insight into their variable perceptions and behaviors, these categories are neither absolute nor static. In other words, liberals and conservatives sometimes align more closely in their perceptions of institutions or subsets of these institutions, and this alignment can increase or decrease over time and across contexts. One 14 example of this phenomenon is the attitude of liberals and conservatives toward larger companies (i.e., “Big Business”). When it comes to larger firms, the distinction between conservatives and liberals vis-à-vis their positivity toward the private sector becomes somewhat blurred, and this “blurring” has been increasing lately (Woolridge 2022). While conservatives generally favor the free market over government solutions, many are nonetheless skeptical (like their liberal counterparts) of big business and the largest companies (Pew 2021). The lukewarm perspective of American conservatives on big business has been long-standing, beginning in the late 19th century when conservative politicians spearheaded antitrust legislation and efforts to dismantle monopolies (Tepper 2019). Somewhat paradoxically, conservatives’ opposition to large, monopolistic business is, in many ways, consistent with their principles vis-à-vis the private sector and the economy. In the extreme, monopolies can stifle the free market and the free exchange of property between individuals, and thus antipathy towards big business can be (and historically has been) framed by conservatives as a pro-free market, pro-private sector policy (Tepper 2019). This phenomenon is likewise observable empirically. For example, examining recent data from the American National Election Study (ANES 2021), which includes a diverse assortment of variables on Americans’ political ideology, political behaviors, and attitudinal measures on a wide range of issues and institutions, illustrates this phenomenon. Expectedly, this data shows a 40-point gap (on a 100-point scale) between liberals and conservatives in “feeling thermometer” ratings when asked about their perceptions of “capitalists” (between very negative for extreme liberals and very positive among extreme conservatives). However, that gap drops to only 27 points when asked about “big business,” with liberals again very negative but conservatives relatively less positive (Figure 2) (ANES 2021). That is, while a substantial majority of extreme 15 liberals have negative feelings about big business, extreme conservatives, while more positive, are still only lukewarm. As noted above, this phenomenon is driven by a long-running and today growing antipathy among many conservatives to larger businesses that they perceive to be, like government, interfering in their free exchanges. Moreover, some have argued that an emerging American “populist right” may result in a “divorce” between conservatism and big business, with conservatives directing outrage at big businesses that are increasingly involved in social issues (Woolridge 2022). Given these dynamics, we propose that the effect of conservatism on customer feedback value is attenuated by the size of the company with which the customer is engaged, with the effect becoming weaker as the size of the company increases. That is, since conservatives’ positivity towards the private sector weakens vis-à-vis the largest companies, we anticipate that the value of their feedback to firms will likewise weaken as firm size increases. Therefore, we hypothesize: H3: The positive relationship between conservatism and customer feedback value is weaker for larger firms. Moderator: External Firm Recognition Product and service modifications, innovations, and related improvements are time- consuming undertakings for firms that often evolve incrementally across more than just one customer purchase occasion (Warren and Sorescu 2017a). In addition, depending on the modality, firms often receive a wide variety of feedback ranging from insights into specific functional attributes (e.g., product delivery time) to more generic suggestions (e.g., launching a new product in a new category). However, not all feedback a firm receives can be incorporated into its offerings. Consequently, most customers neither observe tangible outcomes of their 16 feedback nor have the means to know whether the firm has made a genuine attempt at incorporating their feedback. Indeed, this reality possibly contributes to the low feedback rates that managers encounter today. It also results in an adverse selection problem because the firm has certain information about its intentions and ability to incorporate customer feedback (an “invisible” quality), whereas customers do not have access to such information when they make the decision about whether to provide feedback. Based on the information asymmetry literature (Spence 1973; Spence 2002; Connelly et al. 2011), an effective tool for addressing an adverse selection problem is for the firms with higher levels of the “invisible” quality to generate a credible signal of that quality (Cao et al. 2022). Such signaling allows customers to differentiate the high-quality firms from the low-quality firms (Lampel and Shamsie 2000). We identify external firm recognition, defined as awards and similar recognition received by a firm for its products and services from independent third-party entities, as a signal that can be generated only by those firms that make genuine efforts to listen to customers and incorporate customer feedback to achieve product or service superiority. Examples of such recognition include OAG Aviation Worldwide's “Most On-Time Airline” award and Business Insurance’s “U.S. Insurance Awards for Leadership, Inventiveness, and Ingenuity in Products and Services.” We argue that such external recognition received by firms will strengthen the positive relationship between conservatism and the value of feedback. This is because external recognition received by firms signal their invisible intention and efforts made towards improving products and services (Gemser, Leenders, and Wijnberg 2008). Since conservatives have a higher (relative to liberals) baseline level of trust in firms, they are not only more likely to accept this signal of product and service superiority (c.f., Connelly et al. 2011), but also more likely to use this signal to make inferences regarding the firm’s intention and ability to listen to 17 customers’ voices and incorporate customer feedback to improve its offerings. Therefore, we hypothesize that: H4: The positive relationship between conservatism and customer feedback value is stronger for firms with higher levels of external recognition. 18 METHODS AND RESULTS To test our hypotheses, we use a combination of real-world, large-sample consumer survey data and experimental data. In Study 1, we analyze a large secondary dataset from multiple sources to demonstrate the main effect of customer political ideology on the managerial usefulness of customer feedback, as well as the moderating effects of firm size and external firm recognition. In Study 2a, we utilize a recall study with a product context (as opposed to the service-based context of Study 1) to replicate the main effect of customer political ideology on feedback usefulness and provide evidence for its underlying process through trust in the private sector and in the firm. Finally, in Study 2b, we replicate the relationships observed from Study 2a using a controlled context experiment characterized by a combination of product and service elements. Study 1: Main Effect and Moderation Design and Sample. In Study 1, we examine the main effect of political ideology on the customer feedback value (Hypothesis 1) and the moderating roles of firm size (Hypothesis 3) and external firm recognition (Hypothesis 4). To test these relationships, we compiled a large secondary dataset from three independent sources. First, information about customers' political ideology, their experiences with, and feedback to, the firms from which they have recently purchased, along with their demographic data, was obtained from the American Customer Satisfaction Index (ACSI). The ACSI is a U.S.-based syndicated market research company specializing in measuring customer satisfaction with the products and services offered by hundreds of federal government agencies and over 400 private sector firms across almost 50 consumer industries. For federal government agencies, the ACSI interviews several thousand customers annually and includes political ideology as a relevant segmentation variable within the 19 questionnaire. After completing this federal government services questionnaire, respondents are then randomly screened for additional surveys from a subset of applicable private sector industries. We obtained data on 15,953 customers of 117 firms who completed both the federal government questionnaire and a private sector questionnaire for the ACSI between 2016 and 2020, providing information about customers' political ideology and their experiences with (and feedback to) firms. Second, drawing on prior research, we obtain data on multiple measures of firm size from the Standard and Poor’s (S&P) Compustat Capital IQ and Center for Research in Security Prices (CRSP) databases, including annual sales, total assets, number of employees, market share, and market value of equity. Lastly, we obtained data on the news coverage of each firm, including product and service-related awards received by firms, from the Ravenpack (Edge) News Analytics database. Ravenpack compiles information on a variety of events (e.g., new product releases, product recalls, awards, and many others) for thousands of entities (including companies, people, events, and commodities) from all major news wire and other Internet sources and collates it into structured data using text analytics tools. It has been used in prior research to gather marketing- related information such as new product introductions by firms (Warren and Sorescu 2017a; 2017b, Varma, Bommaraju, and Singh 2023) and inter-market competition (Samadi 2016). Specifically, we obtained Document Sentiment Score (DSS) and Document Sentiment Confidence (DSC) variables of all news items under the “awards” event category for the period from 2015 to 2020 for all the firms included in our sample.3 The “awards” category captures 3DSS represents “the sentiment for the news document by analyzing the language in each and every sentence. The score is calculated by averaging the sentiment across all sentences in the document.” It ranges from -1.00 to +1.00. DSC is “a measure of the confidence when determining the DSS. This confidence score is measured from the same aggregated sentiment distribution used to compute the DSS.” It ranges from 0.00 to +1.00 with values under 0.20 and above 0.80 representing very low confidence and very high confidence, respectively. 20 news covering awards received by firms for their products and services. To ensure that the news items can be ascribed to the focal firm, we use only those with an entity relevance score of 100 (Warren and Sorescu 2017a; 2017b). After compiling data from these multiple sources, our final sample comprises 10,808 cross-sectional customer observations regarding their political ideologies and their experiences with 58 large U.S. firms spanning four service-dominant economic sectors, i.e., commercial airlines, consumer retail, internet services providers, and insurance, for the period from 2016 to 2020.4 Variables and Measurement Dependent Variable (Managerial Usefulness of Feedback). The feedback provided by customers is beneficial if the company can utilize it to improve its products/services or create new products (Kumar et al. 2010; Kumar and Bhagwat 2010). In the ACSI’s annual study, customers have the option of providing written feedback to the company from which they have recently purchased. They can do so through two open-ended survey questions that specifically ask them to provide feedback on what they liked and did not like about the products and services of the company based on their recent experiences. To the best of our knowledge, no prior study measures the managerial usefulness of feedback provided by customers in the form of written text. Therefore, we adopt a multi-step procedure to construct a measure of the managerial customer feedback value: (1) identification of its dimensionality, (2) coding on the identified dimensions, and (3) validation by practicing managers. 4ACSI collects survey data to model the satisfaction of customers across 11 sectors (10 private sector and the federal government sector). As per its data collection protocol, since the federal government sector study is unrelated to the private sector study, the data collection for the latter is randomly combined with that of four economic sectors, namely commercial airlines, consumer retail, internet services providers, and insurance. The complete list of firms measured by ACSI in these sectors and additional details about both ACSI’s annual private sector study and federal government sector study is available on its website (www.theacsi.org). While ACSI’s federal government data has been used extensively in public policy research (e.g., Sharma et al. 2018; Morgeson et al. 2022), our study is the first to use it in the marketing literature. 21 In step (1), since feedback is viewed as akin to knowledge for firms in the customer engagement literature (Kumar and Pansari 2016), we draw on prior studies on consumption knowledge (Clarkson, Janiszewski, and Cinelli 2013) and market knowledge (Prabhu, Chandy, and Ellis 2005) to identify two key dimensions of the usefulness of feedback: (1) shared customer feedback breadth, which refers to the range of topics (e.g., product quality, service quality, price, etc.) covered in customer feedback, and (2) shared customer feedback depth, which refers to the amount of within-topic information covered in customer feedback. To complement this deductive approach that draws on prior literature, we also adopt an inductive grounded theory approach, wherein we interviewed eight product managers on what they consider to be useful customer feedback from the perspective of helping them improve their products and services (Zeithaml et al. 2020). The two dimensions of breadth and depth emerged from their responses as well (for details, see APPENDIX A). In step (2), two coders independently coded 1,200 randomly selected instances of customer feedback within the ACSI sample using a predetermined coding scheme, with an initial inter-coder agreement rate of over 90%, and then reached full coder agreement based on discussions. During the coding, each feedback instance was assigned one of the three levels (scores), i.e., low (1), medium (2), and high (3) on both dimensions. Shared customer feedback breadth and depth were then combined to form the measure of customer feedback value (APPENDIX A, Table 2). This measure takes the value “0” if customers do not provide feedback to the company or provide gibberish, and it takes values from 1 to 6 based on the breadth and depth of the feedback. As such, it is left-censored (or censored from below) and we account for this unique distribution through our modeling strategy (discussed later). In the final step (3), we assess the validity of our measure to ensure that it indeed captures 22 the usefulness of feedback from a managerial perspective. For this, we recruited 200 managers from the Prolific platform for reasonable compensation and had them rate the usefulness of a randomly chosen sample of feedback. Specifically, we randomly (and equally) assigned managers to one of the four conditions (each having 40 feedback instances from commercial airlines, retail, insurance, and internet service provider sectors) and asked them to “rate how useful each customer feedback is for you to improve your product or service in the future” (1 = “Not useful at all”; 6 = “Very useful”). We observe a correlation of 0.75 between our measure of usefulness and managerial coding of usefulness, significantly higher than other recent studies that use a similar correlation-based approach to assess the validity of new measures (e.g., Malshe, Colicev, and Mital 2020, Table 3). Finally, using 1,080 of the 1,200 coded observations, we trained an ordinal classification model in conjunction with a Random Forest learner (Fernández-Delgado et al. 2014; Frankel, Jennings, and Lee 2022) to predict the usefulness of feedback for the rest of the sample and used the remaining 120 for out-of-sample validation of prediction (for more details, see APPENDIX B). The predicted values provide our dependent variable in Study 1. Key Predictor variables (Customer Political Ideology, Firm Size, and External Firm Recognition). Customer political ideology is the primary predictor in our conceptual model and is captured using a single-item scale (Jung and Mittal 2020, 2021; “On the scale below, please indicate which best represents your political identity; 1 = “extremely liberal,” and 7 = “extremely conservative”; M = 3.86, SD = 1.87).5 Table 4 in APPENDIX D provides a summary of the distribution of the sample across political ideology groups for Study 1 (and Study 2a and 2b). 5Drawing on prior political ideology research in marketing, we use this single-item measure of political ideology across all three studies. The extensive use of this measure in research is driven by findings that a single left-right item has sufficient explanatory power to make further dimensions of political ideology redundant (Jost 2006; Jost, Federico, and Napier 2009). 23 Further, drawing on prior research in the marketing-finance interface (Kumar, Peterson, and Leone 2013), we operationalize firm size as the natural log of the total annual sales (M = 10.33, SD = 1.27). The data on annual sales was obtained from the Compustat Capital IQ database. Next, we gather data on external firm recognition from the Ravenpack (Edge) News Analytics database. Our measure of external firm recognition is the sum of the product of DSS and DSC of all “awards” news items of the focal firm from the quarter previous to the one in which an interview of the focal customer was undertaken by the ACSI. Taking a sum of all the news items covering a given award allows us to better capture the media coverage and hence potential exposure to information about such awards by the customers (Stabler and Fischer 2020). The summed measure was then rescaled to -1 to +1 before entering the models to facilitate the interpretation of the estimates. Control variables. Drawing on extant research on customer post-purchase evaluations (e.g., Morgeson et al. 2020) and political identity in marketing (Jung and Mittal 2021; Fernandes et al. 2022), we include a set of relevant controls in our analysis. These include segmentation factors of customer age (continuous; M = 49.98, SD = 16.52), gender (1 = “male”, 0 = “all others”; M = 0.45, SD = 0.50), education (1 = “less than high school”, 5 = “post-graduate”; M = 3.50, SD = 1.04), income (1= “under $20K”,7 = “$100K or more”; M = 4.17, SD = 1.86), and race (1 = “white”, 0 = “others”; M = 0.84, SD = 0.36), along with customer satisfaction (1 = “not at all satisfied”, 10 = “very satisfied”; M = 8.22, SD = 1.91). We also include as controls the normalized sum of sentiment scores (DSS*DSC) of all the positive (continuous; M = 0.08, SD = 0.11) (excluding those from the “awards” category) and negative (continuous; M = -0.5, SD = 0.09) news items concerning the focal firms in our sample for the same period as our moderator of external firm recognition. 24 Model Specification To test Hypotheses 1, 3, and 4, we estimate regression models of customer feedback value as a function of customers’ political ideology, along with the relevant control variables discussed earlier. Further, we include a set of fixed effects to account for potential biases arising from omitted variables. Our dataset spans five years (2016-2020), a period includes a plethora of time- varying factors, such as three changes in the political party-in-power in the U.S. and a significant economic crisis (resulting from the COVID-19 pandemic), among many other unobserved time- varying events of potential significance. Such factors inarguably affected the political discourse in the U.S. as well as firms’ ability to invest in customer experiences (Bloom, Fetcher, and Yeh 2021). Hence, we include time (month-year) fixed effects to account for these and any other time-varying factors of importance omitted from our models (Mittal et al. 2005). We also include firm fixed effects that account for unobserved firm strategies that can potentially influence both customer political ideology (e.g., firm strategies in the political domain like political spending, corporate sociopolitical activism, etc.) and customer experiences (e.g., strategies concerning providing incentives to customers for sharing feedback with firms). While the inclusion of such control variables and fixed effects accounts for potential omitted variable biases in our model to a significant degree (Wooldridge 2010), biases may still be present, resulting in customer political ideology being potentially endogenous. Given this, we adopt a control function approach with appropriate instrumental variables as part of our identification strategy to address any potential remaining omitted variable bias. Drawing on the literature on the effect of political advertising on voters' political dispositions during elections (e.g., Fowler, Franz, and Ridout 2021), we construct two instrumental variables based on the cost of political advertising. Specifically, we use the total cost of political advertisements aired during 25 the past three years (for the year in which the customer was surveyed by the ACSI) at the Designated Marketing Area (DMA) level for both the Democratic and Republican party candidates to construct our instrumental variables of Democratic political advertising intensity and Republican political advertising intensity, respectively. Data obtained from the Wesleyan Media Project (WMP) – Kantar Media/CMAG on more than 2.5 million broadcast TV political advertisements aired during the period 2013-2018 was used to create these instruments.6 Valid instrumental variables should satisfy both the relevance criterion (i.e., they are conceptually correlated with the potentially endogenous political ideology variable) and the exclusion criterion (i.e., they do not directly impact the dependent variable). The relevance of our instrumental variables stems from the well-identified effect of political advertisements on voters' political dispositions. Evidence from both randomized field experiments (Kendall, Nannicini, and Trebbi 2015) and quasi-experiments (Spenkuch and Toniatti 2018) shows a meaningful change in voters’ political dispositions caused by exposure to political advertisements. Concerning their validity, political broadcast advertisements are exogenous to the usefulness of feedback provided by customers to private sector firms. First, the general time trend of political advertisements coincides with elections, which are preset by law. For example, significantly more political advertisements occur during election years (particularly in the few months before the election) compared with non-election years.7 Furthermore, political advertisements are used 6The Wesleyan Media Project tracks all broadcast advertisements aired by or on behalf of federal and state election candidates in every Demographic Market Area (DMA) or media market in the country. In the U.S., almost all the TV advertising is strategized and purchased at the DMA level (Goldstein and Freedman 2002; Spenkuch and Toniati 2018). TV advertising accounts for about 73.4% of the campaign budgets of political parties (Rideout, Fowler, and Franz 2021) and is possibly more effective than advertising on digital media (Coppock, Green, and Poter 2022), making it appropriate to construct our instrumental variables. We use the cost of advertising because it closely reflects actual exposure to advertisements. 7For example, the years 2017 (a non-election year) and 2018 (a midterm election year) saw 122,067 and 1,209,430 TV political advertisement broadcasts in the U.S, respectively. 26 by political parties to appeal for votes or other forms of support. Taken together, it is unlikely that political advertisements directly affect customer behavior toward firms over and above the effects of the focal customer’s political ideology and other control variables and fixed effects in our models. Regarding the diagnostic statistics on the relevance of our instrument variables, the F- value of a regression with the potentially endogenous variable of customer political ideology as the dependent variable and only the instruments as independent variables is significantly higher than 10. Further, in the first-stage regression (Table 5, APPENDIX D), we find that both instruments are statistically significant (Democratic political advertising intensity: p < 0.01; Republican political advertising intensity: p = 0.03) and the F-test value (21.17, p < .01) for their stepwise inclusion is greater than 10 (Staiger and Stock 1994). In the first-stage regression, the instruments also affect customer political ideology in the expected directions, such that Democratic political advertising intensity and Republican political advertising intensity have negative and positive signs, respectively. Since we use two instruments, we conduct the Hansen J statistic of over-identification. The J statistic value of 0.68 is statistically non-significant (p = 0.41), indicating that over-identification is not a concern with our instruments. Overall, these results rule out the weak instrument problem. Our full model specification is as follows: 𝑌𝑖 = 𝛼𝑖 + 𝛽1 𝐶𝑢𝑠𝑡𝑜𝑚𝑒𝑟 𝑃𝑜𝑙𝑖𝑡𝑖𝑐𝑎𝑙 𝐼𝑑𝑒𝑜𝑙𝑜𝑔𝑦𝑖 + 𝛽2 𝐶𝑢𝑠𝑡𝑜𝑚𝑒𝑟 𝑃𝑜𝑙𝑖𝑡𝑖𝑐𝑎𝑙 𝐼𝑑𝑒𝑜𝑙𝑜𝑔𝑦𝑖 × 𝐹𝑖𝑟𝑚 𝑆𝑖𝑧𝑒𝑗 + 𝛽3 𝐶𝑢𝑠𝑡𝑜𝑚𝑒𝑟 𝑃𝑜𝑙𝑖𝑡𝑖𝑐𝑎𝑙 𝐼𝑑𝑒𝑜𝑙𝑜𝑔𝑦𝑖 × 𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐹𝑖𝑟𝑚 𝑅𝑒𝑐𝑜𝑔𝑛𝑖𝑡𝑖𝑜𝑛𝑗 + 𝛽4𝐹𝑖𝑟𝑚 𝑆𝑖𝑧𝑒𝑗 + 𝛽5𝐸𝑥𝑡𝑒𝑟𝑛𝑎𝑙 𝐹𝑖𝑟𝑚 𝑅𝑒𝑐𝑜𝑔𝑛𝑖𝑡𝑖𝑜𝑛𝑗 + 𝛽 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖 + 𝛽 ∑ 𝑌𝑒𝑎𝑟 𝐹𝑖𝑥𝑒𝑑 𝐸𝑓𝑓𝑒𝑐𝑡𝑠𝑡 + 𝛽 ∑ 𝐹𝑖𝑟𝑚 𝐹𝑖𝑥𝑒𝑑 𝐸𝑓𝑓𝑒𝑐𝑡𝑠𝑗 + 𝜀𝑖 (1) where i denotes the customer, j denotes the firm, and the outcome variable Yi is the customer i’s usefulness of feedback provided to firm j. Our parameters of interest are β1, which is the estimate of the main effect of customer political ideology to test Hypothesis 1, and β2 and β3, which are 27 the estimates of the moderating effects of firm size to test Hypothesis 3 and external firm recognition to test Hypothesis 4, respectively. As discussed earlier, our dependent variable of the usefulness of feedback is censored from below (i.e., left-censored) because it takes on a value of 0 for those customers who did not provide feedback with firms and takes on positive values (1 through 6) for those who did. As such, we estimate our model (equation 1) using a Tobit specification, which accounts for such censoring by treating the dependent variable Yi as a latent variable Y*. The dependent variable Yi is observed as 0 when the latent variable Y* ≤ 0 and Y* when Y* > 0. We cluster standard errors at the firm level to address the potential non- independence of observations on customers clustered within any given firm (Abadie et al. 2017).8 Following the standard control function approach (Papies, Ebbes, and van Heerde 2017), we first estimate the first-stage regression model using OLS with customer political ideology as the dependent variable and our instrumental variables as the key predictors, along with all other control variables and fixed effects from equation (1). We then predict the residuals of this model and incorporate them into our full model (equation 1) as the endogeneity correction term (control function) that we estimate using Tobit. We follow the recommended control function approach of simultaneous estimation of both equations with bootstrapping (500 replications) to get correct standard errors, as the control function is an estimated quantity (Papies, Ebbes, and van Heerde 2017). Estimation Results. The results from the instrumental variable Tobit estimations are summarized in Table 1, where column (1) provides estimates from the main effect model, and 8The results are robust to alternate clustering of standard errors by industry and time (month-year). 28 column (2) provides estimates from the full interaction model.9 To begin, the endogeneity correction term (control function) is statistically significant in both models, indicating the presence of endogeneity and that it is being accounted for (Wooldridge 2015). In the main effects model, in support of Hypothesis 1, we find that an increasingly conservative ideology is associated with higher customer feedback value to firms (0.6424, p < 0.01). On average, one-unit increase in ideology towards conservatism is associated with 0.64 increase in the value of feedback. Turning to the interaction model, in support of Hypothesis 3, we find that firm size negatively moderates this relationship such that the effect of an increasingly conservative ideology on the usefulness of feedback is weaker for larger firms (-0.0376, p < 0.01). Lastly, external firm recognition positively moderates the positive effect of political ideology on the usefulness of feedback such that it is stronger for firms with higher levels of recognition received (0.2010, p < 0.01), supporting Hypothesis 4. We graphically plot these interaction effects in Figure 8 (APPENDIX C), showing significantly different slopes for the effect of political ideology on customer feedback value at different levels (two standard deviations above and below the mean) of firm size and external firm recognition. Robustness checks Alternate Proxies of Firm Size. Prior studies in the accounting and finance literature have found that the effects of firm size could differ based on the use of different proxies (Dang, Li, and Yang 2018). Therefore, to assess the robustness of our findings concerning the moderating role of firm size (Hypothesis 3), we alternatively operationalize firm size using four different proxies: the natural log of the total number of employees (M = 4.80, SD = 1.09), total assets (M 9In ancillary models, we also estimate Tobit models without instrumental variables. The results from those models are comparable to the results with instrumental variables and are reported in columns (1)-(2) of Table 7 (APPENDIX D). 29 = 10.19, SD = 1.50), the market value of equity (M = 9.88, SD = 1.91), and market share (M = 0.31, SD = 0.28) (Bhattacharya, Morgan, and Rego 2021; Boyd, Chandy, and Cunha 2010; Dang, Li, and Yang 2018). The data on these four additional proxies of firm size was also obtained from the Compustat Capital IQ database (number of employees and total assets) and the Center for Research on Security Prices (CRSP) (market value of equity). Next, we alternatively operationalize firm size relative to the size of other firms in the industry. Specifically, we operationalize it as market share, computed by dividing the sales of the focal firm by the total sales of all the firms operating in the same industry. High levels of market share are indicative of increasingly monopolistic behaviors by firms, and we predict that conservatives will have a less favorable opinion about such monopolistic firms (Bhattacharya, Morgan, and Rego 2021). We use the NAICS classification to identify other firms operating in each industry and obtain sales data for all firms in such industries from the Compustat Capital IQ database. The results from the control function estimation using these additional proxies, summarized in Table 6 (APPENDIX D), are similar to our main models. Unlike other studies that examine objective implications of firm size and find that the results change based on the use of different measures of firm size (Dang, Li, and Yang 2018), our results concerning firm size are stable, potentially because the effects are driven by consumers’ perception of firm size and not the objective constraints or advantages associated with firm size, as discussed earlier in the manuscript. Alternate Estimation Strategies (Ordered Probit). As discussed earlier, our main models are estimated using Tobit estimation, which accounts for the censored nature of our dependent variable of the usefulness of customer feedback. We now assess the robustness of our findings to an alternate estimation that makes different assumptions concerning our dependent variable. To 30 better account for the ordinal nature of our dependent variable,10 we employ an ordered probit estimation using the standard procedure (e.g., Crolic et al. 2022). Similar to our main models, we adopt a control function approach using the same two instrumental variables. Instead of Tobit, we use ordered probit to estimate our full model (equation 1). The results from this alternate estimation, summarized in Columns (3) – (4) of Table 7 (APPENDIX D), are similar to the main models and continue to provide support for Hypotheses 1, 3, and 4. Enhancing Customer Feedback Analysis through Managerial Evaluation and BERT Technology. To enhance the measurement and prediction of customer feedback value from a managerial perspective, this robustness check utilized the raw ratings provided by managers on 500 instances of customer feedback, which served as the ground truth. Specifically, managers rate the usefulness of each instance of customer feedback on a 5 point-scale (“Please rate each set of feedback below on how useful it is for improving your company's customer experience in the future”; 1 = “not useful at all,” and 7 = “extremely useful”). This strategy is based on the idea that managers' evaluations directly show how useful feedback is to a firm, making them crucial for assessing the true value of customer feedback. To scale this assessment process and enable automated prediction across the entire dataset, we employed the BERT (Bidirectional Encoder Representations from Transformers) model. BERT is a language model introduced by Google in 2018. It is built upon the transformer architecture, a cutting-edge deep learning framework that has revolutionized the field of text analysis. BERT's major strength lies in its ability to understand the context of words in a sentence. Most traditional text analysis techniques such as bag-of-words that I used in the 10The variable is ordinal because our coding scheme, discussed in greater detail in APPENDIX A, assigns increasingly higher categories (low, medium, and high) to the breadth of topics and depth within each topic of the customer feedback. The distance between either pair of adjacent categories is not assumed to be the same. 31 previous section, or LIWC, which has been used a lot in marketing research, treat each word in a sentence independently. BERT considers the entire sequence of words, allowing it to capture sematic nuances in the sentences, resolve ambiguities in words with multiple meanings, and in general achieve higher accuracy in NLP tasks, including text classification, the task at hand here. Additionally, BERT can be fine-tuned for specific contexts, offering remarkable flexibility and adaptability. We applied this new predicted customer feedback value that was trained directly using managerial rating of the usefulness of feedback combined with the enhanced language model to our main effect and interaction models. The results, summarized in Table 8 (APPENDIX D), consistently support Hypotheses 1, 3, and 4. Topics of Feedback from Different Customer Segments. To detect any potential differences in the topics covered by different ideology-based customer segments in their feedback to firms, we conduct an additional analysis using Non-negative Matrix Factorization (NMF), a topic modeling algorithm (Kuang, Choo, and Park 2015).11 Specifically, we build models to investigate the topics covered by all customers together and then separately by conservative, liberal, and moderate customers in their feedback. To ensure an appropriate comparison of feedback among customers who had relatively similar experiences (irrespective of political ideology), we limit this analysis to the consumer retail stores sector, which has the largest sample in our data. For all the customer segments mentioned above, the same three sets of topics emerge from the models, namely price and selection, customer service, and product quality and variety (for more details, see Figure 9, APPENDIX C). This finding of consistency among topics covered by different customer segments indicates that if managers were to reach out to only one customer segment for feedback, it would not result in a topically biased set of 11We thank an anonymous reviewer for encouraging us to explore whether conservative customers’ feedback is representative of overall customers. 32 feedback. Studies 2a/2b: Main Effect and Underlying Process In Study 1, we demonstrated a positive main effect of customer political ideology on customer feedback value, indicating that the more conservative a customer is, the more likely she is to provide useful feedback to the firm (Hypothesis 1). This effect is attenuated for larger firms (Hypothesis 3) and strengthened for firms receiving higher levels of external recognition (Hypothesis 4). In Studies 2a and 2b, we aim to establish the serial mediating role of trust in the private sector and trust in the firm in the relationship between political ideology and feedback value (Hypothesis 2). Specifically, Study 2a tests this relationship and the underlying process with an actual prior purchase in a product context, in contrast to the service-based context of Study 1. Study 2b further establishes the robustness of Study 2a findings with a controlled context of customer experience. Study 2a: Method Participants and Design. One hundred and ninety-two U.S.-based adult participants (Mage = 40, 68% female) completed the study online for payment on the Prolific platform. Participants were asked to recall a recent purchase experience of new clothing made in the past 12 months. They reported the brand and company from which they purchased, and whether they had provided feedback to the company. Those who provided feedback were asked to recall its content; those who did not provide feedback due to lack of opportunity were asked whether they would have provided feedback (if given the chance) and what it would have been. We use the same coding scheme as in Study 1 (APPENDIX A) to manually code the usefulness of feedback in this study. Measures. Participants completed a 7-point item of trust in the company (“Overall, how 33 much do you trust this company?”; 1 = “completely distrust,” and 7 = “completely trust”; M = 5.51, SD = 0.93) and rated their overall customer satisfaction with the purchase using a 7-point scale (1 = “extremely dissatisfied,” and 7 = extremely satisfied”; M = 6.27, SD = 0.89). They then rated their political ideology on a 7-point scale (“On the scale below, please indicate which best represents your political identity”; 1 = “extremely liberal,” and 7 = “extremely conservative”; M = 3.67, SD = 1.93), trust in the private sector on a 5-point scale (“How much do you trust private businesses to do what is right?”; 1 = “Never,” and 5 = “Always”; M = 2.71, SD = .93), and several demographic questions, including gender, age, ethnicity, education, and household income. Study 2a: Results Customer Feedback Value. Similar to Study 1, two coders independently coded each customer feedback instance and reached a coder agreement of about 85%. Discrepancies were resolved through discussion. We then use a Tobit regression model where feedback value is a function of customer political ideology, controlling for customer satisfaction and demographic variables (gender, age, ethnicity, education, and household income). Replicating Study 1’s main effect for Hypothesis 1, we find a positive association between political ideology and the feedback value (β = .64, p = .001), with conservatives providing more valuable feedback to firms. Mediation analysis. To test the underlying process (Hypothesis 2), we conduct a serial mediation (Model 6, Hayes 2017) with customer political ideology as the independent variable, trust in the private sector and trust in the firm as the serial mediators (in that order), and the feedback value as the dependent variable. The results reveal a significant indirect effect of customer political ideology on the feedback value via trust in the private sector and trust in the 34 firm (β = .0120, SE = .0068, 95%CI = [.0024, .0280]; Figure 3, APPENDIX C). Study 2b: Method Participants and Design. Two hundred and three adult participants (Mage = 40, 57% female) completed the study online for payment on the Prolific platform. Participants read a scenario of purchasing a new carpet from the fictitious home improvement retailer, Home Project. They were shown a survey invitation from Home Project asking for their feedback about the purchase and installation experience of the new carpet (see APPENDIX C, Figure 5-7 for details of the research stimuli). Participants were then asked whether they would like to provide feedback, and if so, what feedback they would provide. We use the same coding scheme to manually code the feedback value as in Study 1 and Study 2a (APPENDIX A). Measure. In Study 2b, we adopt a different measure of trust in the firm, using a 7-point, 4-item scale (Morgan and Hunt 1994; Ganesan 1994). Sample items included “Home Project has high integrity” and “Home Project can be counted on to do what is right.” We measure customer political ideology, trust in the private sector, and demographic questions using the same scales as in Study 2a.12 Study 2b: Results Customer Feedback Value. Following the same process of coding and modeling used in Study 1 and Study 2a, we replicate the main effect concerning Hypothesis 1 and find a positive association between political ideology and the customer feedback value (β = .22, p = .043), with conservatives providing more valuable feedback to firms. Mediation analysis. We conduct the same serial mediation (Model 6, Hayes 2017) 12Since it is a controlled experiment and all of the respondents had the “same experience”, we do not measure customer satisfaction as a control variable in Study 2b. We continue to include all other controls variables from Study 2a in Study 2b. 35 deployed in Study 2a and replicate the findings of the indirect effect of customer political ideology on the feedback value via trust in the private sector and trust in the firm (β = .0396, SE = .0151, 95%CI = [.0156, .0735]; Figure 4, APPENDIX C). Discussion. Using a recall study and a controlled context experiment in Study 2a and 2b, respectively, we confirm Hypothesis 1 that conservative ideology is associated with a higher level of feedback value provided by customers to firms. More importantly, Study 2a and Study 2b support the role of trust as the mechanism underlying the relationship between customer political ideology and the customer feedback value. These studies demonstrate that conservative consumers have higher trust in the private sector, which further helps brands and companies build trust with customers along the customer journey, eventually leading customers to provide more valuable feedback (Hypothesis 2). Additionally, Studies 2a and 2b tested the robustness of the relationship between consumer political ideology and the customer feedback value in the context of consumer products and a combination of product and service elements, respectively, compared with services in Study 1. 36 DISCUSSION AND FUTURE RESEARCH General Discussion Today, firms face significant challenges in collecting useful customer feedback, including low response rates, poor quality of feedback, and feedback with limited detail, making it inadequate or less valuable for managerial decision-making (Forbes 2022). Despite substantial investments in soliciting customer feedback (Statista 2022), these challenges persist. Our research highlights important findings about a novel driver of customer feedback value – political ideology, the process underlying its effect, and managerially relevant boundary conditions (firm size and external recognition) (Figure 1, APPENDIX C). Drawing on conservatives' shared belief in the free market, the private sector, and the capitalist model of society, we theorize that conservatives, compared to liberals, have higher trust in private sector institutions. Such heightened trust toward the private sector facilitates conservatives’ trust-building process with the firms, leading them to share more valuable feedback with firms. However, this positive effect of political conservatism on customer feedback value weaker for larger firms. This is because as the size of the firm approaches the level of a “big business,” conservatives tend to view these firms less favorably – as “big business” can be seen to stifle the free market – narrowing the gap between conservatives' and liberals’ trust in these firms. Further, we show that external third-party recognition received by the firm strengthens the effect of political conservatism on customer feedback value, reinforcing conservatives’ higher trust in firms. These findings make important contributions to both research and practice. Implications for Research and Theory Our research extends the literature on customer feedback by identifying an observable 37 and inherent customer characteristic – customer political ideology – as a key driver of customer feedback behavior. In doing so, we also complement existing findings on the political ideology– customer complaint behavior relationship from Jung et al. (2017) and introduce an objective measure of customer feedback value that captures the usefulness of customer feedback from a managerial perspective. First, prior literature on customer feedback behaviors has predominantly focused on firm- level or customer-level outcomes of soliciting or providing feedback (e.g., Bone et al. 2017; Celuch, Robinson, and Walsh 2015; Challagalla, Venkatesh, and Kohli 2009), but has been largely silent about the antecedents of customer feedback behavior. We illustrate that customer political ideology plays an important role in driving customer feedback behavior, such that customers that hold a stronger conservative ideology exhibit a higher likelihood to provide valuable feedback to firms. This is because conservative customers tend to have higher baseline trust in firms, which is rooted in their positive view of the private sector and their belief in the free market. Our findings concerning the role of customer trust in driving customer feedback behavior align with the commitment-trust theory of relationship marketing (Morgan and Hunt 1994). Investment of time and energy to provide feedback to a firm is an act of cooperation that reflects customers’ commitment to the exchange relationship, which is driven by customers’ trust in the firm (Morgan and Hunt 1994). At a broader level, our findings suggest that beyond firm-level factors, customers’ characteristics and dispositions play a crucial role in driving customer engagement, particularly those that encourage customers to forge stronger exchange relationships. Future research could investigate factors such as customers’ independent vs. interdependent cultural backgrounds (Ma, Yang, and Mourali 2014) and moral dispositions (which can also be related to political ideology; 38 Graham, Haidt, and Nosek 2009) as antecedents of customer engagement behavior. Another customer disposition that future research may consider is customers’ implicit theories, which are customers’ beliefs about whether the nature of human characteristics, institutions, and the world can be changed or not. Different customer implicit theories may lead to different levels of motivation to provide feedback considering one of the main purposes of feedback is to help an agency to make changes to improve products or services (Dweck, Chiu, and Hong 1995). In the current research, we account for the potential role of the firm’s political ideology in shaping customer behavior by including firm fixed effects (Study 1) and using a controlled context experiment (Study 2b). Future research could investigate how the interplay of customer’s and firm’s political ideology influences customer trust and customer feedback behavior, especially in the context of firms' participation in sociopolitical issues, such as corporate social responsibility, brand activism, and corporate political activity (e.g., Bhagwat et al. 2020; Hydock, Paharia, and Blair 2020). Second, in their examination of customer complaints to third-party regulatory institutions, Jung et al. (2017) demonstrate that conservative consumers are less likely than liberal consumers to file complaints against firms. Our research complements these findings by demonstrating that conservative customers are in fact more likely to provide valuable feedback, both positive and negative, directly to firms. The findings from the two works combined indicate that conservatives do not necessarily differ from liberals when it comes to voicing their dissatisfaction with the firms, but instead that they may choose different channels (third-party regulatory agencies vs. directly to the firm) to do so. Underlying such different preferences for channels, as we argue, is the difference in the underlying motivation to complain, and more importantly, different levels of trust and commitment in the exchange relationships with firms. 39 Third, drawing on the extant knowledge management literature and based on interviews with industry experts, we develop a behavioral measure that captures the extent to which specific customer feedback is useful in helping managers develop or improve their products and services. In this way, our research complements earlier work by Kumar and Pansari (2017) which developed a self-reporting-based measure of customer knowledge value, opening additional avenues for future research to examine actual customer feedback value. We also provide evidence for the validity of the feedback value measure in the form of a high correlation (.75) between usefulness scores based on this new measure and those by practicing managers on a randomly selected sample of customer feedback. This measure provides a foundational step in understanding customer feedback usefulness and facilitating the future development and testing of theory about the antecedents and outcomes of customer feedback usefulness (Haws, Sample, and Hulland 2023). This article also makes two important contributions to political ideology research in marketing (Jung and Mittal 2020). While recent studies have answered calls for research on the role of political ideology in shaping broader customer behaviors that are not explicitly political in nature (Jung and Mittal 2017), most of them identify such behaviors as traceable to either the epistemic motives or existential motives that underlie the motivational substructure of political ideology (Jung and Mittal 2020; Jost 2009). We extend this stream of literature by building on a third class of psychological factor that forms the motivational sub-structure of political ideology – relational motives comprised of needs for a shared reality, political socialization, and group justification (Jost 2009). We demonstrate that the need for a shared reality, as manifest in conservatives’ (vs. liberals’) shared belief in the most efficient organization of society by means of an unfettered free market or private sector (vs. the government sector), drives them to place 40 higher trust in firms (Pew 2019) and subsequently to provide more valuable feedback to firms as a form of cooperation (Morgan and Hunt 1994). Additional attitudinal or behavioral outcomes of political ideology that potentially result from the other two relational motives, i.e., political socialization and group justification, remain underexplored and can inform future research. For example, such research might investigate the role of political ideology in driving customer polarization in response to firm actions (e.g., price increases, product withdrawals, change in brand imagery) (Kübler et al. 2020; Luo, Weis, and Raithel 2013) and their preference (or aversion) to brands that exemplify shared experiences and engagement in brand communities such as Jeep, Lego, and Harley Davidson (Herhausen et al. 2019). In addition, we answer a recent call for research to “investigate when the effect of political ideology on consumer behaviors is strong or weak rather than simply examining whether political identity affects consumer behavior” (Jung and Mittal 2021, p. 574). We do so by identifying a context that challenges (firm size) and a context that reinforces (external firm recognition) the trust-based mechanisms underlying the effect of consumer conservatism on feedback behaviors as moderators. Our core assertion for the effect of consumer conservatism on feedback value is that it stems from conservatives’ (vs. liberals’) belief in the most efficient organization of society by means of an unfettered free market or private sector (vs. the government sector). However, as the size of private firms becomes larger, they tend to appear more monopolistic, thereby reducing the difference between conservative and liberal customers’ trust placed in these firms. As a result, larger firms do not benefit as much from conservatives’ greater proclivity to provide valuable feedback. While several studies in marketing examine the objective implications of firm size (e.g., Marinova 2004; Bhattacharya, Morgan, and Rego 2021), this is the first study, and particularly within the blossoming political ideology research in 41 marketing, to demonstrate a unique perception-based challenge posed by firm size from the customer perspective. Furthermore, through our examination of external firm recognition as a moderator that strengthens the effect of customers’ conservatism on feedback value, we offer insights that suggest how firms can further enhance the value of feedback from conservative customers. We explicate the presence of information asymmetry between firms and customers when it comes to feedback behaviors and delineate the role of external firm recognition as signals that resolve such asymmetry (Spence 1973; Spence 2002; Connelly et al. 2011). Specifically, external recognition received by firms for their products and services act to resolve the adverse selection problem because they signal the invisible firm intention and efforts made towards improving products and services. Given their baseline higher trust, conservative customers are more receptive of this signal of product and service superiority and are more likely to use this signal to make inferences regarding the firm’s intention and ability to incorporate customer feedback in their decisions. Such an approach to signaling, as a tool from a firm's perspective that is instrumental in addressing adverse selection, is new to the political ideology literature in marketing (Jung and Mittal 2020). Managerial Implications For firms, the development of innovative new offerings and/or the enhancement of current offerings that fulfill ever-evolving consumer needs is imperative to stay competitive (Currim, Lim, and Kim 2012; Kumar and Pansari 2017; Rubera and Kirca 2012). This is evidenced in the case of Apple, which generated $25 million in additional annual revenue by actively seeking and responding to customer feedback. While customer intelligence is a key input to this iterative process, firms often struggle to solicit it (Markey, Rechheld, and Dullweber 42 2009). In this research, we discover a large and easily identifiable segment of customers (Jung et al. 2017) – those with a conservative political ideology – that firms can leverage for more valuable feedback, as these customers are more likely to have higher trust in firms and thus more likely to provide useful feedback. As such, practitioners can use our findings to enhance the effectiveness of their market research spending towards listening to the voice of the customer. Further, knowing the contexts wherein political ideology is a weaker or stronger predictor of customer feedback usefulness is crucial for managerial decision-making concerning knowledge- seeking from customers. To that end, our findings show that firms that receive more external recognition may benefit more from conservative customers’ tendency to provide valuable feedback, but that larger firms may not benefit as much as smaller firms. In ancillary analysis, we demonstrate that soliciting customer feedback from politically conservative customers does not result in a biased set of feedback since customers across the political spectrum (i.e., liberals, moderates, and conservatives) provide feedback on similar topics. Nonetheless, we interpret our findings cautiously and emphasize that there are still numerous benefits of soliciting feedback from all customers, including liberals and moderates. Several studies from prior literature demonstrate such positive customer-level outcomes of firms’ feedback-soliciting behaviors and customers’ feedback-sharing behaviors. For example, in the service industry, the act of employees seeking customer feedback may be viewed as a customer- oriented behavior that increases customers’ social benefit perceptions and strengthens customer- firm exchange relationships (Celuch, Robinson, and Walsh 2015); recalling a positive experience in the feedback process increases a customer’s future purchases (Bone et al. 2017); and feedback seeking may increase customer satisfaction (Challagalla, Venkatesh, and Kohli 2009). Firms, however, should also avoid soliciting feedback too frequently, as repeatedly 43 soliciting feedback may have detrimental cumulative effects on customers’ future purchase frequency and the amount spent (Flynn, Salisbury, and Seiders 2017; Dholakia, Singh, Westbrook 2010). Firm-initiated customer engagement behaviors such as feedback solicitation and stimulating word-of-mouth can lead to increased risk perceptions among shareholders, decreasing the market value of the firm (Beckers, van Doorn, and Verhoef 2018). We suggest that managers use our findings in their decision-calculus in combination with contextual factors such as the scenarios and purposes of the feedback they require at hand, and the establishment of the product/service, the brand, or the firm. For example, managers can consider reaching out to conservative customers if they need to collect detailed feedback in a short period of time; similarly, managers of less established products or brands can also reach out to conservative customers for feedback at the early stages following market launch since conservative customers tend to have higher baseline trust towards an unfamiliar private sector entity. Finally, our research also provides managers with direct guidance on how to analyze potentially large-scale, unstructured customer feedback data and utilize customer knowledge efficiently. We demonstrate that managers can manually code a relatively small sample of customer feedback in terms of the usefulness of the feedback to improve their products or services, and then utilize NLP and a simple classification model to predict the potential value of each “piece” of customer feedback they collect. In this way, managers can channel their energies to those customers’ feedback with the highest value. Further, topic modeling techniques can help managers easily gather all feedback that pertains to a specific topic of managerial interest. Managers can also easily monitor the trend of customer feedback value of different geographical regions and evaluate the effectiveness of any firm-initiated customer engagement campaigns from the perspective of receiving high-quality customer feedback. 44 REFERENCES Abadie, A., Athey, S., Imbens, G. W., & Wooldridge, J. (2017). When should you adjust standard errors for clustering? (No. w24003). National Bureau of Economic Research. American National Election Studies. 2021. ANES 2020 Time Series Study Full Release [dataset and documentation]. July 19, 2021 version. www.electionstudies.org. Aksoy, L., Banda, S., Harmeling, C., Keiningham, T. L., & Pansari, A. (2022). Marketing’s role in multi-stakeholder engagement. International Journal of Research in Marketing, 39(2), 445-461. Alterman, E. (2013). The Cause: The Fight for American Liberalism from Franklin Roosevelt to Barack Obama. Penguin. Anderson, J. C., & Narus, J. A. (1990). A model of distributor firm and manufacturer firm working partnerships. Journal of Marketing, 54(1), 42-58. Bailey, J. R., & Phillips, H. (2020). How do consumers feel when companies get political. Harvard Business Review. https://hbr. org/2020/02/how-do-consumers-feel-when- companies-get-political. Beckers, S. F., Van Doorn, J., & Verhoef, P. C. (2018). Good, better, engaged? The effect of company-initiated customer engagement behavior on shareholder value. Journal of the Academy of Marketing Science, 46, 366-383. Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity, and social history. Games and Economic Behavior, 10(1), 122-142. Berlin, I. (2017). Two concepts of liberty. In The Liberty Reader (pp. 33-57). Routledge. Berry, L. L., & Parasuraman, A. (1991). Marketing Services (New York. NY: Free Press. Bhagwat, Y., Warren, N. L., Beck, J. T., & Watson IV, G. F. (2020). Corporate sociopolitical activism and firm value. Journal of Marketing, 84(5), 1-21. Bhattacharya, C. B., Rao, H., & Glynn, M. A. (1995). Understanding the bond of identification: An investigation of its correlates among art museum members. Journal of Marketing, 59(4), 46-57. Bhattacharya, C. B., & Sen, S. (2003). Consumer–company identification: A framework for understanding consumers’ relationships with companies. Journal of Marketing, 67(2), 76- 88. Bhattacharya, A., Morgan, N. A., & Rego, L. L. (2021). Customer satisfaction and firm profits in monopolies: a study of utilities. Journal of Marketing Research, 58(1), 202-222. 45 Bloom, N., Fletcher, R. S., & Yeh, E. (2021). The impact of COVID-19 on US firms (No. w28314). National Bureau of Economic Research. Bone, S. A., Lemon, K. N., Voorhees, C. M., Liljenquist, K. A., Fombelle, P. W., Detienne, K. B., & Money, R. B. (2017). “Mere measurement plus”: how solicitation of open-ended positive feedback influences customer purchase behavior. Journal of Marketing Research, 54(1), 156-170. Boyd, D. E., Chandy, R. K., & Cunha Jr, M. (2010). When do chief marketing officers affect firm value? A customer power explanation. Journal of Marketing Research, 47(6), 1162- 1176. Cao, Z., Song, R., Sorescu, A., & Chua, A. (2022). Innovation Potential, Insider Sales, and IPO Performance: How Firms Can Mitigate the Negative Effect of Insider Selling. Journal of Marketing. Celuch, K., Robinson, N. M., & Walsh, A. M. (2015). A framework for encouraging retail customer feedback. Journal of Services Marketing. Challagalla, G., Venkatesh, R., & Kohli, A. K. (2009). Proactive postsales service: When and why does it pay off?. Journal of Marketing, 73(2), 70-87. Chen, J., Xu, A. J., Rodas, M. A., & Liu, X. (2023). Order matters: Rating service professionals first reduces tipping amount. Journal of Marketing, 87(1), 81-96. Chesbrouh, H. W., & Appleyard, M. M. (2007). Open innovation and strategy. California Management Review, 50(1), 57-76. Choma, B. L., Hanoch, Y., Gummerum, M., & Hodson, G. (2013). Relations between risk perceptions and socio-political ideology are domain-and ideology-dependent. Personality and Individual Differences, 54(1), 29-34. Clarkson, J. J., Janiszewski, C., & Cinelli, M. D. (2013). The desire for consumption knowledge. Journal of Consumer Research, 39(6), 1313-1329. Crolic, C., Thomaz, F., Hadi, R., & Stephen, A. T. (2022). Blame the bot: anthropomorphism and anger in customer–chatbot interactions. Journal of Marketing, 86(1), 132-148. Cooper, R. G., & Kleinschmidt, E. J. (1995). Benchmarking the firm's critical success factors in new product development. Journal of Product Innovation Management: An International Publication of the Product Development & Management Association, 12(5), 374-391. Coppock, A., Green, D. P., & Porter, E. (2022). Does digital advertising affect vote choice? Evidence from a randomized field experiment. Research & Politics, 9(1), 20531680221076901. 46 Connelly, B. L., Certo, S. T., Ireland, R. D., & Reutzel, C. R. (2011). Signaling theory: A review and assessment. Journal of Management, 37(1), 39-67. Currim, I. S., Lim, J., & Kim, J. W. (2012). You get what you pay for: the effect of top executives’ compensation on advertising and R&D spending decisions and stock market return. Journal of Marketing, 76(5), 33-48. Dang, C., Li, Z. F., & Yang, C. (2018). Measuring firm size in empirical corporate finance. Journal of Banking & Finance, 86, 159-176. de Oliveira Santini, F., Ladeira, W. J., Pinto, D. C., Herter, M. M., Sampaio, C. H., & Babin, B. J. (2020). Customer engagement in social media: a framework and meta-analysis. Journal of the Academy of Marketing Science, 48(6), 1211-1228. Del Rio, A. B., Vazquez, R., & Iglesias, V. (2001). The effects of brand associations on consumer response. Journal of Consumer Marketing. Devos, T., Spini, D., & Schwartz, S. H. (2002). Conflicts among human values and trust in institutions. British Journal of Social Psychology, 41(4), 481-494. Dholakia, U. M., Singh, S. S., & Westbrook, R. A. (2010). Understanding the effects of post- service experience surveys on delay and acceleration of customer purchasing behavior: Evidence from the automotive services industry. Journal of Service Research, 13(4), 362- 378. Dick, A. S., & Basu, K. (1994). Customer loyalty: toward an integrated conceptual framework. Journal of the Academy of Marketing Science, 22(2), 99-113. Dose, D. B., Walsh, G., Beatty, S. E., & Elsner, R. (2019). Unintended reward costs: The effectiveness of customer referral reward programs for innovative products and services. Journal of the Academy of Marketing Science, 47(3), 438-459. Duggan, Maeve and Aaron Smith (2016), “The Political Environment on Social Media,” Pew Research Center (October 6), https://www.pewresearch.org/internet/2016/10/25/the- political-environmenton-social-media// Dweck, C. S., Chiu, C. Y., & Hong, Y. Y. (1995). Implicit theories and their role in judgments and reactions: A word from two perspectives. Psychological Inquiry, 6(4), 267-285. Erikson, R. S., & Tedin, K. L. (2015). American public opinion: Its origins, content and impact. Routledge. Escalas, J. E., & Bettman, J. R. (2003). You are what they eat: The influence of reference groups on consumers’ connections to brands. Journal of Consumer Psychology, 13(3), 339-348. Escalas, J. E., & Bettman, J. R. (2009). Connecting with celebrities: Celebrity endorsement, 47 brand meaning, and self-brand connections. Journal of Marketing Research, 13(3), 339- 348. Everett, J. A. (2013). The 12 item social and economic conservatism scale (SECS). PloS one, 8(12), e82131. Farmer, B. R. (2005). American conservatism: History, theory and practice. Cambridge Scholars Press. Fernandes, D., Ordabayeva, N., Han, K., Jung, J., & Mittal, V. (2022). How political identity shapes customer satisfaction. Journal of Marketing, 00222429211057508. Fernández-Delgado, M., Cernadas, E., Barro, S., & Amorim, D. (2014). Do we need hundreds of classifiers to solve real world classification problems?. The Journal of Machine Learning Research, 15(1), 3133-3181. Flynn, A. G., Salisbury, L. C., & Seiders, K. (2017). Tell us again, how satisfied are you? The influence of recurring posttransaction surveys on purchase behavior. Journal of Service Research, 20(3), 292-305. Frank, E., & Hall, M. (2001, September). A simple approach to ordinal classification. In European Conference on Machine Learning (pp. 145-156). Springer, Berlin, Heidelberg. Frankel, R., Jennings, J., & Lee, J. (2022). Disclosure sentiment: Machine learning vs. dictionary methods. Management Science, 68(7), 5514-5532. Forbes. (2022). Customer Feedback Analysis: How Your Customers Help You Improve Your Business. Retrived online at: https://www.forbes.com/sites/forbestechcouncil/2022/12/06/customer-feedback-analysis- how-your-customers-help-you-improve-your-business/?sh=2f59cfc84727. Forbes. (2011). Another Myth Bites The Dust: How Apple Listens To Its Customers. Retrived online at: https://www.forbes.com/sites/stevedenning/2011/08/26/another-myth-bites-the- dust-how-apple-listens-to-its-customers/?sh=2e9aeee0cde5. Fortune. (2010). Inside the Secret World of Trader Joe’s. Retrived online at: https://fortune.com/2010/08/23/inside-the-secret-world-of-trader-joes/. Fournier, S. (1998). Consumers and their brands: Developing relationship theory in consumer research. Journal of Consumer Research, 24(4), 343-373. Fowler, E. F., Franz, M., & Ridout, T. N. (2021). Political advertising in the United States. Routledge. Ganesan, S. (1994). Determinants of long-term orientation in buyer-seller relationships. Journal 48 of Marketing, 58(2), 1-19. Gemser, G., Leenders, M. A., & Wijnberg, N. M. (2008). Why some awards are more effective signals of quality than others: A study of movie awards. Journal of Management, 34(1), 25-54. Goenka, S., & Van Osselaer, S. M. (2021). Why Is It Wrong to Sell Your Body? Understanding Liberals’ Versus Conservatives’ Moral Objections to Bodily Markets. Journal of Marketing, 00222429211046936. Goldstein, K., & Freedman, P. (2002). Campaign advertising and voter turnout: New evidence for a stimulation effect. Journal of Politics, 64(3), 721-740. Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conservatives rely on different sets of moral foundations. Journal of Personality and Social Psychology, 96(5), 1029. Han, K., Jung, J., Mittal, V., Zyung, J. D., & Adam, H. (2019). Political identity and financial risk taking: Insights from social dominance orientation. Journal of Marketing Research, 56(4), 581-601. Haws, K. L., Sample, K. L., & Hulland, J. (2023). Scale use and abuse: Towards best practices in the deployment of scales. Journal of Consumer Psychology, 33(1), 226-243. Hayes, A. F. (2017). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford publications. Herhausen, D., Ludwig, S., Grewal, D., Wulf, J., & Schoegel, M. (2019). Detecting, preventing, and mitigating online firestorms in brand communities. Journal of Marketing, 83(3), 1- 21. Harmeling, C. M., Moffett, J. W., Arnold, M. J., & Carlson, B. D. (2017). Toward a theory of customer engagement marketing. Journal of the Academy of Marketing Science, 45(3), 312-335. Hekman, D. R., Aquino, K., Owens, B. P., Mitchell, T. R., Schilpzand, P., & Leavitt, K. (2010). An examination of whether and how racial and gender biases influence customer satisfaction. Academy of Management Journal, 53(2), 238-264. Ho-Dac, N. N., Carson, S. J., & Moore, W. L. (2013). The effects of positive and negative online customer reviews: do brand strength and category maturity matter?. Journal of Marketing, 77(6), 37-53. Howison, J. D. (2018). The Historical Origins and Contemporary Dynamics of Conservatism in the United States: Anticommunism, the New Class Critique, and the Environment. Political Studies Review, 16(1), 13-24. 49 Hydock, C., Paharia, N., & Blair, S. (2020). Should your brand pick a side? How market share determines the impact of corporate political advocacy. Journal of Marketing Research, 57(6), 1135-1151. Islam, J. U., Hollebeek, L. D., Rahman, Z., Khan, I., & Rasool, A. (2019). Customer engagement in the service context: An empirical investigation of the construct, its antecedents and consequences. Journal of Retailing and Consumer Services, 50, 277-285. Jaakkola, E., & Alexander, M. (2014). The role of customer engagement behavior in value co- creation: a service system perspective. Journal of Service Research, 17(3), 247-261. Joshi, A. W., & Sharma, S. (2004). Customer knowledge development: antecedents and impact on new product performance. Journal of Marketing, 68(4), 47-59. Jost, J. T., (2006). The End of the End of Ideology. American Psychologist, 61 (7), 651–70. Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism as motivated social cognition. Psychological Bulletin, 129(3), 339. Jost, J. T., Federico, C. M., & Napier, J. L. (2009). Political ideology: Its structure, functions, and elective affinities. Annual Review of Psychology, 60, 307-337. Jost, J. T., Nosek, B. A., & Gosling, S. D. (2008) Its Resurgence in Social, Personality, and Political Psychology. Jost, J. T., Federico, C. M., & Napier, J. L. (2009). Political ideology: Its structure, functions, and elective affinities. Annual Review of Psychology, 60(1), 307-337. Jost, J. T. (2017). Ideological asymmetries and the essence of political psychology. Political Psychology, 38(2), 167-208. Jung, K., Garbarino, E., Briley, D. A., & Wynhausen, J. (2017). Blue and red voices: Effects of political ideology on consumers’ complaining and disputing behavior. Journal of Consumer Research, 44(3), 477-499. Jung, J., & Mittal, V. (2020). Political identity and the consumer journey: A research review. Journal of Retailing, 96(1), 55-73. Jung, J., & Mittal, V. (2021). Political Identity and Preference for Supplemental Educational Programs. Journal of Marketing Research, 58(3), 559-578. Kendall, C., Nannicini, T., & Trebbi, F. (2015). How do voters respond to information? Evidence from a randomized campaign. American Economic Review, 105(1), 322-53. Khamitov, M., Wang, X., & Thomson, M. (2019). How well do consumer-brand relationships drive customer brand loyalty? Generalizations from a meta-analysis of brand relationship 50 elasticities. Journal of Consumer Research, 46(3), 435-459. Khan, R., Misra, K., & Singh, V. (2013). Ideology and brand consumption. Psychological Science, 24(3), 326-333. Kidwell, B., Farmer, A., & Hardesty, D. M. (2013). Getting liberals and conservatives to go green: Political ideology and congruent appeals. Journal of Consumer Research, 40(2), 350-367. Kim, C. H., Han, D., Duhachek, A., & Tormala, Z. L. (2018). Political identity, preference, and persuasion. Social Influence, 13(4), 177-191. Konya-Baumbach, E., Schuhmacher, M. C., Kuester, S., & Kuharev, V. (2019). Making a first impression as a start-up: Strategies to overcome low initial trust perceptions in digital innovation adoption. International Journal of Research in Marketing, 36(3), 385-399. Kramer, R. M. (1994). The sinister attribution error: Paranoid cognition and collective distrust in organizations. Motivation and Emotion, 18, 199-230. Kuang, D., Choo, J., & Park, H. (2015). Nonnegative matrix factorization for interactive topic modeling and document clustering. Partitional Clustering Algorithms, 215-243. Kumar, V., & Bhagwat, Y. (2010). Listen to the customer. Marketing Research, 22(2), 14-19. Kumar, V., Aksoy, L., Donkers, B., Venkatesan, R., Wiesel, T., & Tillmanns, S. (2010). Undervalued or overvalued customers: Capturing total customer engagement value. Journal of Service Research, 13(3), 297-310. Kumar, V., & Pansari, A. (2016). Competitive advantage through engagement. Journal of Marketing Research, 53(4), 497-514. Kumar, V., Petersen, J. A., & Leone, R. P. (2010). Driving profitability by encouraging customer referrals: Who, when, and how. Journal of Marketing, 74(5), 1-17. Kumar, V., Rajan, B., Gupta, S., & Pozza, I. D. (2019). Customer engagement in service. Journal of the Academy of Marketing Science, 47(1), 138-160. Kübler, R. V., Langmaack, M., Albers, S., & Hoyer, W. D. (2020). The impact of value-related crises on price and product-performance elasticities. Journal of the Academy of Marketing Science, 48, 776-794. Lacey, R., Suh, J., & Morgan, R. M. (2007). Differential effects of preferential treatment levels on relational outcomes. Journal of Service Research, 9(3), 241-256. Lardieri, Alexa (2020). Trump’s Approval Drops as Partisan Divide Hits All-Time High, Poll Finds. USNews.com (July 6). Retrived online at: https://bit.ly/3ndFNb1. 51 Lam, S. K., Ahearne, M., Hu, Y., & Schillewaert, N. (2010). Resistance to brand switching when a radically new brand is introduced: A social identity theory perspective. Journal of Marketing, 74(6), 128-146. Lampel, J., & Shamsie, J. (2000). Critical push: Strategies for creating momentum in the motion picture industry. Journal of Management, 26(2), 233-257. Levi, M., & Stoker, L. (2000). Political trust and trustworthiness. Annual Review of Political Science, 3(1), 475-507. Locke, J. (2013). Two Treatises of government, 1689. The anthropology of citizenship: A reader, 43-46. Lovelock, C., Wirtz, J., & Bansal, H. S. (2008). Services marketing, Canadian edition. Luo, X., Wiles, M., & Raithel, S. (2013). Make the most of a polarizing brand. Harvard Business Review, (NOV). Ma, Z., Yang, Z., & Mourali, M. (2014). Consumer adoption of new products: Independent versus interdependent self-perspectives. Journal of Marketing, 78(2), 101-117. Macpherson, C. B. (1951). Locke on capitalist appropriation. Western Political Quarterly, 4(4), 550-566. Malshe, A., Colicev, A., & Mittal, V. (2020). How main street drives wall street: Customer (dis) satisfaction, short sellers, and abnormal returns. Journal of Marketing Research, 57(6), 1055-1075. Marinova, D. (2004). Actualizing innovation effort: the impact of market knowledge diffusion in a dynamic system of competition. Journal of Marketing, 68(3), 1-20. Markey, R., Reichheld, F., & Dullweber, A. (2009). Closing the customer feedback loop. Harvard Business Review, 87(12), 43-47. McKnight, D. H., & Chervany, N. L. (2006). Reflections on an initial trust-building model. Handbook of Trust Research, 29. Microsoft. (2017). State of Global Customer Service Report. https://info.microsoft.com/rs/157- GQE-382/images/EN-CNTNT-Report-DynService-2017-global-state-customer- service.pdf. Mittal, V., Anderson, E. W., Sayrak, A., & Tadikamalla, P. (2005), “Dual Emphasis and the Long-Term Financial Impact of Customer Satisfaction,” Marketing Science, 24(4), 544- 555. Moorman, C., & Day, G. S. (2016). Organizing for marketing excellence. Journal of 52 Marketing, 80(6), 6-35. Moorman, C., Zaltman, G., & Deshpande, R. (1992). Relationships between providers and users of market research: The dynamics of trust within and between organizations. Journal of Marketing Research, 29(3), 314-328. Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58(3), 20-38. Morgeson III, F. V., Hult, G. T. M., Mithas, S., Keiningham, T., & Fornell, C. (2020). Turning complaining customers into loyal customers: Moderators of the complaint handling– Customer loyalty relationship. Journal of Marketing, 84(5), 79-99. Morgeson III, F. V., Sharma, P. N., Sharma, U., & Hult, G. T. M. (2022). Partisan bias and citizen satisfaction, confidence, and trust in the US Federal Government. Public Management Review, 24(12), 1933-1956. Narveson, Jan. The libertarian idea. Broadview Press, 2001. Nilsson, A., & Jost, J. T. (2020). The authoritarian-conservatism nexus. Current Opinion in Behavioral Sciences, 34, 148-154. Niskanen, W. A., & Cato Institute (Washington). (1988). Reaganomics: An insider's account of the policies and the people (Vol. 292). New York: Oxford University Press. Ode, K. “Are consumer rating surveys driving you nuts? You're not alone — survey fatigue is scoring a 10,” StarTribune.com. October 29, 2016. Retrieved online at: https://www.startribune.com/are-consumer-rating-surveys-driving-you-nuts-you-re-not- alone-survey-fatigue-is-scoring-a-10/398948031/. Ordabayeva, N., & Fernandes, D. (2018). Better or different? How political ideology shapes preferences for differentiation in the social hierarchy. Journal of Consumer Research, 45(2), 227-250. Paharia, N., & Swaminathan, V. (2019). Who is wary of user design? The role of power-distance beliefs in preference for user-designed products. Journal of Marketing, 83(3), 91-107. Palmatier, R. W., Houston, M. B., Dant, R. P., & Grewal, D. (2013). Relationship velocity: Toward a theory of relationship dynamics. Journal of Marketing, 77(1), 13-30. Pansari, A., & Kumar, V. (2017). Customer engagement: the construct, antecedents, and consequences. Journal of the Academy of Marketing Science, 45(3), 294-311. Park, C. W., MacInnis, D. J., Priester, J., Eisingerich, A. B., & Iacobucci, D. (2010). Brand attachment and brand attitude strength: Conceptual and empirical differentiation of two critical brand equity drivers. Journal of Marketing, 74(6), 1-17. 53 Papies, D., Ebbes, P., & Heerde, H. J. V. (2017). Addressing endogeneity in marketing models. In Advanced Methods for Modeling Markets (pp. 581-627). Springer, Cham. Pew (2019). In a Politically Polarized Era, Sharp Divides in Both Partisan Coalitions. pewresearch.org. Retrieved online at https://www.pewresearch.org/politics/2019/12/17/views-of-government-and-the-nation/. Pew (2021). Beyond Red vs. Blue: The Political Typology. pewresearch.org. Retrived online at: https://www.pewresearch.org/politics/2021/11/09/beyond-red-vs-blue-the-political- typology-2/. Prabhu, J. C., Chandy, R. K., & Ellis, M. E. (2005). The impact of acquisitions on innovation: poison pill, placebo, or tonic?. Journal of Marketing, 69(1), 114-130. Randhawa, K., Wilden, R., & Hohberger, J. (2016). A bibliometric review of open innovation: Setting a research agenda. Journal of Product Innovation Management, 33(6), 750-772. Reeves, Richard V. “What real liberalism looks like,” Brookings.edu. January 16, 2018. Retreived online at: https://www.brookings.edu/blog/up-front/2018/01/16/what-real- liberalism-looks-like/. Ridout, T. N., Fowler, E. F., & Franz, M. M. (2021, March). Spending fast and furious: Political advertising in 2020. In The Forum (Vol. 18, No. 4, pp. 465-492). De Gruyter. Rubera, G., & Kirca, A. H. (2012). Firm innovativeness and its performance outcomes: A meta- analytic review and theoretical integration. Journal of Marketing, 76(3), 130-147. Samadi, Mehrdad (2016), “Intermarket Competition: Evidence from Trading Venue Short Sales,” working paper, University of North Carolina at Chapel-Hill. Scanlon, S. (2015). Ronald Reagan and the conservative movement. A Companion to Ronald Reagan, 583-607. Sharma, P. N., Morgeson III, F. V., Mithas, S., & Aljazzaf, S. (2018). An empirical and comparative analysis of E-government performance measurement models: Model selection via explanation, prediction, and parsimony. Government Information Quarterly, 35(4), 515-535. Selnes, F., & Sallis, J. (2003). Promoting relationship learning. Journal of Marketing, 67(3), 80- 95. Shimul, A. S. (2022). Brand attachment: a review and future research. Journal of Brand Management, 1-20. Schwartz, S. H., & Boehnke, K. (2004). Evaluating the structure of human values with confirmatory factor analysis. Journal of Research in Personality, 38(3), 230-255. 54 Siegel, R. (2018). Did your favorite company donate to Democratic or Republican campaigns? This startup will tell you. The Washington Post. Retrived online at: https://www.washingtonpost.com/business/2018/12/19/did-your-favorite-company- donate-democratic-or-republican-campaigns-this-startup-will-tell-you/. Statista (2022). Market research industry - statistics & facts. Retrieved online at: https://www.statista.com/topics/1293/market-research/#topicOverview. Stäbler, S., & Fischer, M. (2020). When does corporate social irresponsibility become news? Evidence from more than 1,000 brand transgressions across five countries. Journal of Marketing, 84(3), 46-67. Solak, N., Jost, J. T., Sümer, N., & Clore, G. L. (2012). Rage against the machine: The case for system‐level emotions. Social and Personality Psychology Compass, 6(9), 674-690. Spence, M. 1973. Job market signaling. Quarterly Journal of Economics, 87: 355-374. Spence, M. 2002. Signaling in retrospect and the informational structure of markets. American Economic Review, 92: 434-459. Spenkuch, J. L., & Toniatti, D. (2018). Political advertising and election results. The Quarterly Journal of Economics, 133(4), 1981-2036. Staiger, D. O., & Stock, J. H. (1994). Instrumental variables regression with weak instruments. Stokburger-Sauer, N., Ratneshwar, S., & Sen, S. (2012). Drivers of consumer–brand identification. International Journal of Research in Marketing, 29(4), 406-418. Storey, C., & Larbig, C. (2018). Absorbing customer knowledge: how customer involvement enables service design success. Journal of Service Research, 21(1), 101-118. Tepper, J. (2019). The Conservative Case for Antitrust. Retrieved online at: https://www.theamericanconservative.com/the-conservative-case-for-antitrust-jonathan- tepper/. Ulver, S., & Laurell, C. (2020). Political ideology in consumer resistance: Analyzing far-right opposition to multicultural marketing. Journal of Public Policy & Marketing, 39(4), 477- 493. Umashankar, N., Ward, M. K., & Dahl, D. W. (2017). The benefit of becoming friends: Complaining after service failures leads customers with strong ties to increase loyalty. Journal of Marketing, 81(6), 79-98. Van Doorn, J., Lemon, K. N., Mittal, V., Nass, S., Pick, D., Pirner, P., & Verhoef, P. C. (2010). Customer engagement behavior: Theoretical foundations and research directions. Journal of Service Research, 13(3), 253-266. 55 Varma, R., Bommaraju, R., & Singh, S. S. (2023). EXPRESS: Female Chief Marketing Officers: When and Why Their Marketing Decisions Differ From Their Male Counterparts?. Journal of Marketing Research, 00222437231156902. Vivek, S. D., Beatty, S. E., & Morgan, R. M. (2012). Customer engagement: Exploring customer relationships beyond purchase. Journal of Marketing Theory and Practice, 20(2), 122- 146. Warren, N. L., & Sorescu, A. (2017a). When 1+ 1> 2: How investors react to new product releases announced concurrently with other corporate news. Journal of Marketing, 81(2), 64-82. Warren, N. L., & Sorescu, A. (2017b). Interpreting the stock returns to new product announcements: How the past shapes investors’ expectations of the future. Journal of Marketing Research, 54(5), 799-815. Winterich, K. P., Zhang, Y., & Mittal, V. (2012). How political identity and charity positioning increase donations: Insights from moral foundations theory. International Journal of Research in Marketing, 29(4), 346-354. Woolridge, A. (2022). Big Business and Conservatives Are Headed for Divorce. Bloomberg.com. Wooldridge, J. M. (2010). Econometric Analysis of Cross Section and Panel Data. MIT press. Wooldridge, J. M. (2015). Control function methods in applied econometrics. Journal of Human Resources, 50(2), 420-445. Wu, C., Che, H., Chan, T. Y., & Lu, X. (2015). The economic value of online reviews. Marketing Science, 34(5), 739-754. Zeithaml, V. A., Jaworski, B. J., Kohli, A. K., Tuli, K. R., Ulaga, W., & Zaltman, G. (2020). A theories-in-use approach to building marketing theory. Journal of Marketing, 84(1), 32- 51. 56 APPENDIX A: CUSTOMER FEEDBACK VALUE CODING SCHEME In line with the literature on consumption knowledge (Clarkson, Janiszewski, and Cinelli 2013) and market knowledge (Prabhu, Chandy, and Ellis 2005), we developed a coding scheme for the customer feedback value variable that consists of two dimensions: shared customer feedback breadth and shared customer feedback depth. Shared customer feedback breadth refers to the range of topics covered by customer feedback, such as product quality, service quality, price, availability, variety, etc., and shared customer feedback depth refers to the within field information provided by the customer. These two dimensions also emerged in our interviews with eight managers from different industries (manufacturing, telecommunications and information, healthcare, retail, transportation, and energy utilities). Responding to the question, “From a manager's perspective, how would you define useful customer feedback such that it can help your company improve its products and/or services?”, most managers stressed the importance of shared customer feedback depth (e.g., “Explicit and detailed explanation that can help business pinpoint or replicate issues,” “Feedback that is actionable -- that is, feedback that is specific enough to identify the root cause of the pain point or delight”). The importance of breadth of information in the feedback also emerged from responses such as, “one that clearly states what the problem is, less emotions more information, so that I can circle back with the relevant team to make the customer experience better in the future.” These managers gave an average importance of 4.38/5 to shared customer feedback breadth and of 5/5 to the importance of shared customer feedback depth. As such, we coded shared customer feedback breadth into three levels: low, medium, and high. Low feedback breadth indicates that the feedback doesn’t provide any specific fields (e.g., “I love this brand!”); medium feedback breadth indicates that the feedback covers one or two 57 fields (e.g., “Salesperson is well informed”; “A wide variety of consumer merchandise at competitive prices”); and high feedback breadth indicates that the feedback covers more than two fields (e.g., “Store is neat and clean and I can use the app while in the store to find what I want. Really happy with Target but wouldn’t mind more grocery variety.”) Following the same approach as for breadth, we coded shared customer feedback depth into three levels: low, medium, and high. Low feedback depth indicates that the feedback provides no details about any fields (e.g., “The service is excellent”); medium feedback depth indicates that the feedback provides some descriptive details about at least one field (e.g., “(I like it that) store personnel greet the customer, offers assistance, and then doesn’t impose on the customer again…lets you shop…”); high feedback depth indicates that the feedback provides actionable details that can help improve business (e.g., “A variety of high-quality goods at reasonable prices. Need 10 items or less checkout option.”). We combine the shared customer feedback breadth and depth into the final customer feedback value variable. Based on our interviews with managers discussed earlier, shared customer feedback depth is more important for customer feedback value to firms. Therefore, we combine the two dimensions based on the level of depth first followed by the level of breadth. Specifically, low breadth-low depth, medium breadth-low depth, and high breadth-low depth are assigned values of 1, 2, and 3, respectively, since low-depth feedback has the lowest value for firms for improving their products and services. Medium breadth-medium depth is assigned the value 4, high breadth-medium depth and medium breadth-high depth are assigned the value 5, and high breadth-high depth is assigned the value 6.13 Finally, if customers provide no feedback 13Two categories of low breadth-medium depth and low breadth-high depth are not feasible based on our coding scheme and hence are not included in the table above. This is because customer feedback is assigned to a “low (or 1)” category on breadth when it does not include any specific field. Absence of any field in the customer feedback precludes the possibility of depth within the field. 58 or gibberish feedback, customer feedback value is assigned a value of 0, i.e., as having no value for the firm. Table 2 below provides the coding scheme for customer feedback value based on the shared customer feedback breadth and depth and related examples. This coding scheme was used in both Study 1, Study 2a, and Study 2b. 59 APPENDIX B: CUSTOMER FEEDBACK VALUE PREDICTION Unlike Study 2, where we manually code customer feedback value for each piece of customer feedback, Study 1 has a sample of 15,953 customer feedback for 117 large U.S. firms spanning five service industries from 2016 to 2020. The coding of customer feedback value for Study 1 involves a multistep process. First, we randomly selected 1,200 customer feedback from our dataset and manually coded the customer feedback value based on the coding scheme in APPENDIX A, Table 2. Second, we next implemented a Natural Language Progressing (NLP) algorithm on the full sample of customer feedback data. We first cleaned the customer feedback content in the following steps: 1) removing punctuations, 2) transforming capital letters into lower cases, 3) splitting customer feedback into words, 4) stemming each word into its root form (e.g., “loved” into “love”) and removing stop words (e.g., “the”, “a”, and “I”), and 5) joining the cleaned and stemmed words back together into customer feedback. Then, we create the bag-of-words model by transforming all customer feedback content into a sparse matrix, which consists of 15,953 rows and 6,563 columns. Each row represents one observation of customer feedback, and each column represents a unique stemmed word used in all customer feedback. Each value in the sparse matrix represents the frequency of one stemmed word used in one customer feedback. Since customer feedback value is an ordinal variable that takes values of 0 – 6, we train an ordinal classifier model that takes into account the ordering information in class attributes (Fernández-Delgado et al. 2014; Frank and Hall 2001) and apply it in conjunction with a Random Forest learner. We combine the sparse matrix of customer feedback and industry dummy variables as the predictors because customer feedback value might vary across industries (i.e., useful or actionable customer feedback might be associated with different topics for 60 different industries). We randomly selected 1,080 coded sample for training and the remaining 120 for validation. The accuracy of our classifier model is 63.33%. As shown in the Confusion Matrix in Table 3 below, 84.17% predictions are within ±1 error. 61 APPENDIX C: FIGURES Figure 1. Blurring of the Difference Between Liberals and Conservatives' Perceptions Towards Capitalists and Big Businesses 62 Figure 2. Conceptual Framework 63 Figure 3. Study 2a – Results of Mediation Analyses 64 Figure 4. Study 2b – Results of Mediation Analyses 65 Figure 5. Study 2b – Stimuli Page 1 66 Figure 6. Study 2b – Stimuli Page 2 67 Figure 7. Study 2b – Stimuli Page 3 68 Figure 8. Study 1 – Interaction Plots Notes: Small (Low) and Large (High) indicate values two standard deviations above and below the mean of moderator variables, respectively. 69 Figure 9. Study 1 – Topic Modeling Results: Representative Words Within Topics 70 APPENDIX D: TABLES Table 1. Study 1 – Results Tobit Estimation (Control Function Approach) Variables Political Ideology Political Ideology X External Recognition a Political Ideology X Firm Size a External Recognition Firm Size Positive News Negative News Satisfaction Age Education Income Gender - Male Race - White Endogeneity Correction (control function) Constant Firm Fixed Effects Year-Month Fixed Effects Observations (Customers) (1) Main Effects Model 0.6424*** (0.1828) -- -- -0.1294 (0.1529) 0.0744 (0.4177) 0.1835 (0.3498) -0.3630 (0.3474) -0.0206* (0.0124) 0.0150*** (0.0039) 0.1123*** (0.0396) -0.0784*** (0.0139) -0.6088*** (0.0505) -0.1024 (0.0833) -0.5894*** (0.1831) 3.0830 (6.5485) Yes Yes 10,808 (2) Interaction Model 0.6281*** (0.1724) 0.2010*** (0.0560) -0.0376*** (0.0079) -0.1332 (0.1667) 0.1436 (0.4301) 0.2363 (0.3682) -0.4169 (0.3378) -0.0207* (0.0122) 0.0151*** (0.0036) 0.0987*** (0.0369) -0.0545*** (0.0121) -0.6079*** (0.0535) -0.1135 (0.0790) -0.5782*** (0.1725) 3.7368 (4.7740) Yes Yes 10,808 *** p<0.01, ** p<0.05, * p<0.1 Cluster (by firms) robust standard errors in parentheses. a Political Ideology, External Recognition and Firm Size were mean-centered before creating interaction terms. 71 Customer Feedback Value 0 1 2 3 4 5 6 Table 2. Customer Feedback Value Coding Scheme and Example Breadth Depth Example -- Low -- (Did not provide feedback OR provided gibberish) Low “I love this brand!” Medium Low “The service is excellent” High Low “Great service, amazing personnel, great variety” Medium Medium “Store personnel greet the customer, offer assistance, and then don’t bother the customer again…let you shop…” High Medium “Ulta usually carries a wide variety of brands, so I don't have to go to several stores. Their website is user-friendly, and the shipping is quick. The employees are always helpful and friendly. I really can't think of anything they should improve upon.” Medium High “I can count on them to have the products I need. Many times, the workers don't appear to really like their jobs very much. It is sometimes difficult to find someone to talk to.” High High "They hire people who seem to really care about other people and are able to convey compassion. When one calls for assistance despite the representative's best efforts the software program that guides them through service appears to be very un-manageable and cumbersome. The succession of screens doesn’t seem to help in pinpointing the exact problem. It doesn't seem to be very intuitive as a product support system." 72 Table 3. Confusion Matrix Predicted Label 1 7 0 0 0 0 0 2 2 19 1 7 6 0 3 0 0 0 0 0 0 4 0 2 0 9 5 0 5 0 3 0 8 41 10 6 0 0 0 0 0 0 1 2 3 4 5 6 l e b a L e u r T 73 Political Ideology 1 Extremely Liberal 2 Liberal 3 Slightly Liberal 4 Moderate or Middle of the Road 5 Slightly Conservative 6 Conservative 7 Extremely Conservative Table 4. Distribution of Political Ideology Across Studies Study 2a (%) Study 1 (%) 16.15 13.09 20.83 16.64 10.94 8.85 14.58 25.07 12.50 9.74 19.27 18.69 5.73 7.93 Study 2b (%) 12.32 17.73 11.82 12.81 9.85 24.63 10.84 74 Table 5. Study 1 – First Stage Regression Results of the 2SLS Estimation (1) Main Effects Model -2.25e-08*** (3.98e-09) 1.54e-08** (6.85e-09) 0.1336 (0.1520) 0.2213 (0.3040) 0.1204 (0.2185) 0.0508 (0.1535) -0.0277** (0.0129) 0.0194*** (0.0014) -0.1669*** (0.0211) -0.0067 (0.0132) 0.1436*** (0.0492) 0.2666*** (0.0470) 3.2427** (1.5883) Yes Yes 10,806 Variables Democratic political advertising intensity Republican political advertising intensity External Recognition Firm Size Positive News Negative News Satisfaction Age Education Income Gender - Male Race - White Constant Firm Fixed Effects Year-Month Fixed Effects Observations (Customers) *** p<0.01, ** p<0.05, * p<0.1 Cluster (by firms) robust standard errors in parentheses. 75 Variables Political Ideology Political Ideology X External Recognition a Political Ideology X Firm Size a External Recognition Firm Size Positive News Negative News Satisfaction Age Education Income Gender - Male Race - White Endogeneity Correction (control function) Constant Table 6. Study 1 – Alternate Proxies of Firm Size Tobit Estimation (Control Function Approach) - Interaction Models (4) (1) (3) Market Value Total Assets 0.6052*** (0.1675) 0.2198*** (0.0586) -0.0272*** (0.0069) -0.1341 (0.1630) 0.2889 (0.1902) 0.2221 (0.3658) -0.4393 (0.3394) -0.0228* (0.0117) 0.0156*** (0.0035) 0.0986*** (0.0360) -0.0549*** (0.0119) -0.6028*** (0.0525) -0.1143 (0.0790) -0.5576*** (0.1678) 0.1456 (0.9758) Yes Yes 10,867 (2) Number of Employees 0.6089*** (0.1681) 0.1534*** (0.0556) -0.0338*** (0.0089) -0.1500 (0.1658) 0.4411 (0.4227) 0.2269 (0.3692) -0.4215 (0.3368) -0.0226* (0.0121) 0.0155*** (0.0035) 0.0987*** (0.0360) -0.0549*** (0.0119) -0.6047*** (0.0525) -0.1138 (0.0780) -0.5584*** (0.1683) -0.5110 (1.0337) Yes Yes 10,858 of Equity Market Share 0.6272*** 0.6842*** (0.1865) (0.1819) 0.1299** 0.2035*** (0.0591) -0.0229*** (0.0054) -0.1217 (0.1597) -0.0170 (0.0680) 0.1657 (0.3593) -0.3667 (0.3148) -0.0201 (0.0123) 0.0141*** (0.0040) 0.1226*** (0.0381) -0.0779*** (0.0136) -0.6059*** (0.0496) -0.1144 (0.0852) -0.6340*** (0.0560) -0.0931*** (0.0349) -0.1237 (0.1685) -0.1392 (0.7503) 0.1810 (0.3791) -0.3568 (0.3249) -0.0201 (0.0130) 0.0152*** (0.0038) 0.1120*** (0.0387) -0.0794*** (0.0141) -0.5993*** (0.0520) -0.1005 (0.0821) -0.5764*** (0.1864) 0.1067 (1.4592) Yes Yes 10,765 (0.1833) -0.1060 (1.0335) Yes Yes 10,912 Firm Fixed Effects Year-Month Fixed Effects Observations (Customers) *** p<0.01, ** p<0.05, * p<0.1 Cluster (by firms) robust standard errors in parentheses. a Political Ideology, External Recognition and Firm Size were mean-centered before creating interaction terms. 76 Table 7. Study 1 – Alternate Estimations (1) Main Effects Model 0.0556*** (0.0100) -- Tobit Estimation (2) Interaction Model 0.0556*** (0.0110) 0.2038*** Ordered Probit Estimation (Control Function) (4) Interaction Model 0.3333*** (0.0980) (3) Main Effects Model 0.3277*** (0.0977) Variables Political Ideology Political Ideology X External Recognition a Political Ideology X Firm Size a -- External Recognition Firm Size Positive News Negative News Satisfaction Age Education Income Gender - Male Race - White -0.0586 (0.1363) 0.1962 (0.3861) 0.2757 (0.3066) -0.3432 (0.2751) -0.0366*** (0.0100) 0.0264*** (0.0012) 0.0119 (0.0198) -0.0860*** (0.0117) -0.5304*** (0.0384) 0.0692 (0.0512) (0.0660) -0.0365*** (0.0100) -0.0470 (0.0777) 0.2932* (0.1610) -0.3472* (0.2073) 0.2202 (0.3122) -0.0354** (0.0148) 0.0264*** (0.0021) 0.0133 (0.0190) -0.0855*** (0.0123) -0.5267*** (0.0589) 0.0715 (0.0555) Endogeneity Correction (control function) -- -- -- -- -0.0878 (0.0926) 0.0307 (0.2460) 0.1261 (0.2074) -0.2457 (0.2090) -0.0141** (0.0070) 0.0097*** (0.0021) 0.0533** (0.0214) -0.0317*** (0.0071) -0.3431*** (0.0310) -0.0456 (0.0452) -0.3010*** (0.0978) -- 0.1129*** (0.0322) -0.0204*** (0.0048) -0.0792 (0.0952) 0.0484 (0.2509) 0.1318 (0.2106) -0.2425 (0.2072) -0.0132* (0.0071) 0.0096*** (0.0021) 0.0554** (0.0216) -0.0316*** (0.0072) -0.3424*** (0.0312) -0.0460 (0.0454) -0.3072*** (0.0981) -- Constant Firm Fixed Effects Year-Month Fixed Effects Observations (Customers) b *** p<0.01, ** p<0.05, * p<0.1 Cluster (by firms) robust standard errors in parentheses. a Political Ideology, External Recognition and Firm Size were mean-centered before creating interaction terms. Yes Yes 10,808 Yes Yes 10,808 3.6302 (4.2574) Yes Yes 10,806 5.7704*** (0.5959) Yes Yes 10,806 77 Table 8. Study 1 – Alternate Measure of Customer Feedback Value Tobit Estimation (Control Function Approach) Variables Political Ideology Political Ideology X External Recognition a Political Ideology X Firm Size a External Recognition Firm Size Positive News Negative News Satisfaction Age Education Income Gender - Male Race - White Endogeneity Correction (control function) Constant Firm Fixed Effects Year-Month Fixed Effects Observations (Customers) (1) Main Effects Model 0.2761*** (0.1058) -- -- -0.0510 (0.0903) 0.2639 (0.2663) 0.0041 (0.2220) -0.0246 (0.1959) -0.0232*** (0.0078) 0.0100*** (0.0022) 0.0657*** (0.0244) -0.0440*** (0.0076) -0.3352*** (0.0307) -0.0956* (0.0501) -0.2433** (0.1063) -1.7880 (3.6188) Yes Yes 10,806 (2) Interaction Model 0.2808*** (0.1060) 0.1148*** (0.0357) -0.0198*** (0.0051) -0.0450 (0.0926) 0.2776 (0.2687) 0.0131 (0.2241) -0.0273 (0.1940) -0.0224*** (0.0078) 0.0099*** (0.0022) 0.0674*** (0.0244) -0.0438*** (0.0076) -0.3338*** (0.0307) -0.0957* (0.0504) -0.2479** (0.1064) 0.8377 (2.1758) Yes Yes 10,808 *** p<0.01, ** p<0.05, * p<0.1 Cluster (by firms) robust standard errors in parentheses. a Political Ideology, External Recognition and Firm Size were mean-centered before creating interaction terms. 78