Abstract
The primary focus of brand equity research has been on how brand knowledge creates value for firms through customer behavior in product markets. Using archival data and five experiments, this article tests a framework that outlines the unique role brands play in the labor market. The framework distinguishes between vertical and horizontal differentiation and shows that vertical brand differentiation is associated with lower pay, whereas horizontal brand differentiation is associated with higher pay. Employees are also vertically and horizontally differentiated, and firms high in horizontal brand differentiation pay more for employees who match their brands’ differentiating characteristics (i.e., brand-relevant complementarities). Results show that these brand–pay relationships have important downstream effects on employee behavior and, consequently, on firm profits. Specifically, leveraging vertical brand differentiation to lower pay represents a false economy because profits are attenuated by negative effects on employee productivity and retention. In contrast, when managers at firms high on horizontal brand differentiation pay more, profits increase via the same mediating employee behaviors. Six firm strategies and investments that influence firm bargaining power in the employee–brand matching process are found to moderate the brand–pay relationship and downstream effects on profits.
Keywords
Brand equity is the value a brand contributes to the firm (Farquhar 1989). The primary focus of brand equity research has been on how brand knowledge creates value for firms through customer behavior in product markets, or what Keller (1993) calls “customer-based brand equity.” This includes customer-level outcomes such as willingness to pay (e.g., Datta, Ailawadi, and Van Heerde 2017) and firm-level outcomes such as stock valuations and profits (e.g., Mizik 2014).
Research has also identified a second source of brand equity that affects labor markets. Employee-based brand equity is defined as the “value a brand provides to a firm through its effects on the attitudes and behaviors of its employees” (Tavassoli, Sorescu, and Chandy 2014, p. 677). To date, it has been documented via a single financial outcome: employees’ willingness to accept lower pay (Cable and Turban 2003; DelVecchio et al. 2007; Tavassoli, Sorescu, and Chandy 2014). We contribute to this nascent literature in three ways.
First, we challenge the current understanding of employee-based brand equity and propose that different dimensions of brand knowledge—vertical and horizontal brand differentiation (Dommer, Swaminathan, and Ahluwalia 2013; Ordabayeva and Fernandes 2018; Spiller and Belogolova 2017)—have opposite effects on pay. We show that vertical brand differentiation is associated with lower pay, whereas horizontal brand differentiation is associated with higher pay.
Second, drawing on human capital theory (Becker 1964), we suggest that employees also vary in terms of vertical and horizontal differentiation. Leveraging this distinction, we find that firms 1 high in horizontal brand differentiation pay more to hire employees who match their brand's differentiating characteristics—a like-with-like dynamic known as positive assortative matching—whereas vertically differentiated brands do not systematically match with vertically differentiated employees. Across all types of employee–brand matches, we find that the brand–pay relationship is moderated by a set of firm strategies and investments that influence firm bargaining power across different stages of the matching process. These stages include the firm's demand for labor relative to its supply (labor availability), approach to forming a consideration set of job candidates (labor identification), attractiveness to job candidates beyond pay (labor attraction), and improvements in match quality through training (labor development).
Third, previous research suggests that if brand equity is leveraged to lower pay, it should translate into higher profits (Tavassoli, Sorescu, and Chandy 2014). We challenge this view and show that it represents a false economy. When managers leverage vertical brand differentiation to lower pay, it negatively affects profits due to the mediating effects of lower employee productivity and retention. This means that profits are lower than would have been achieved by the positive effect of brand alone. In contrast, we find that when managers pay more due to horizontal brand differentiation, it increases profits via the same mediating employee behaviors.
The next section describes our theory and predictions about employee–brand matching and its effect on pay, employee behaviors, and profits. This is followed by our empirical strategy, which involves the use of archival data and experiments to test our predictions.
Brands in the Labor Market
What role do brands play in the labor market? To answer this question, we characterize the heterogeneous labor market participants—firms and employees—along dimensions that allow us to describe strategic interactions unique to employee-based brand equity.
Heterogeneous Labor Market Participants
Brand differentiation
Formalized in economics by Hotelling’s (1929) model of market competition and subsequently in Lancaster’s (1979) theory of consumer choice, among others, Tirole (1988, pp. 96–97) specifies that with vertical differentiation, “all consumers agree over the most preferred mix of characteristics and, more generally, over the preference ordering,” while with horizontal differentiation “the optimal choice [at equal prices] depends on the particular consumer.” As outlined by Spiller and Belogolova (2017, p. 970), this distinction is central to marketing: “Marketers and researchers alike typically regard products as differentiated by quality (modeled via vertical differentiation) or taste (modeled via horizontal differentiation),” with market segments (e.g., Desai 2001), products (e.g., Liu, McFerran, and Haws 2020), customer reviews (e.g., Lee, Bollinger, and Staelin 2023), and product lines (e.g., Balachander and Stock 2009) characterized along these two dimensions.
Brands can also be characterized along vertical and horizontal dimensions (Dommer, Swaminathan, and Ahluwalia 2013; Ordabayeva and Fernandes 2018). These dimensions reflect associations based on product (e.g., perceived quality) and nonproduct (e.g., symbolic benefits) attributes (Keller 1993). In line with this literature, we define vertical brand differentiation as the holistic perception of a brand's superior quality relative to other brands and horizontal brand differentiation as the holistic perception of a brand's uniqueness relative to other brands.
Vertically differentiated brands therefore tend to perform well on dimensions most consumers regard positively, such as Vanguard's reliable investment returns or Ritz-Carlton's high level of service. Horizontally differentiated brands, in contrast, perform well on distinctive qualities that may only be valued by some consumers, such as Jeep's rugged image or Dove's natural-beauty positioning. Brands can, of course, be differentiated along both vertical and horizontal dimensions. For example, Dior and Gucci are both vertically differentiated in terms of their universally valued quality of craftsmanship and heritage of excellence as well as horizontally differentiated in terms of Dior's timeless style and classic feminine beauty and Gucci's fashion-forward androgyny, for which preferences diverge as a matter of taste.
Literature in consumer research finds that consumers leverage these vertical and horizontal brand associations. Specifically, consumers seek to transfer these brand meanings to vertically differentiate and signal that they are better than others or to horizontally differentiate and signal their uniqueness (Dommer, Swaminathan, and Ahluwalia 2013; Ordabayeva and Fernandes 2018).
Employee differentiation
We suggest that employees can also be characterized as being vertically and horizontally differentiated. They can be considered vertically differentiated through what Becker (1964) refers to as “general human capital”—universally valued employee knowledge, skills, and traits, such as negotiation and planning skills or a strong work ethic. Some universally valued human capital may be tied to specialized job skills (e.g., coding; Gibbons and Waldman 2004) or occupations (Mayer, Somaya, and Williamson 2012) and has been examined in terms of “person–job fit” in the organizational psychology literature (Kristof-Brown 2000). In line with this literature, we view employees as being more vertically differentiated when their general human capital can be objectively ranked and is valued by many firms.
In contrast, employees can be horizontally differentiated through what Becker (1964) refers to as “specific human capital,” which is not better or worse in an absolute sense, but is subjectively valued by some firms. It is often viewed as the degree of “person–organization fit” based on an applicant's personality, attitudes, or life experiences (Kristof-Brown 2000).
Person–organization fit, referred to in our framework as employee–brand fit, is a highly relevant form of specific human capital given that consumers use their experiences with employees to create and update their brand knowledge (Sirianni et al. 2013). For this reason, a horizontally differentiated brand such as Wildfang, a feminist, socially progressive purveyor of gender-fluid apparel, values female workers who share its brand's masculine/androgynous identity (Mathwick 2017). This characteristic may be negatively valued by Kate Spade, a brand defined by its feminine style. Research on elite professional service firms has likewise observed that firms evaluate client-facing employee fit based on “distinct personalities, derived from the typical extracurricular interests and self-presentation styles” (Rivera 2012, p. 1007). Horizontal differentiation extends to employees working behind the scenes. To that end, Procter & Gamble seeks employees passionate about babies and aims for “everyone working on Pampers to live the brand ideal in everything they [do]” (Stengel 2011, p. 180).
Employee–Brand Matching
Firms and employees match in labor markets with the aim of maximizing the joint value created and shared between them (Becker 1964). Firms define “match quality” in terms of the performance gains employees provide (Weller et al. 2019), while employees focus on the pecuniary and nonpecuniary benefits received for their labor (Campbell, Coff, and Kryscynski 2012; Lucas 1977). For firms to secure strong matches, they need to compensate in accordance with the value employees provide (e.g., Eckert et al. 2022).
Importantly, matches vary in their complementarity—sometimes referred to as production or resource complementarities—meaning certain types of workers create more value at certain types of firms (Becker 1964). For example, faculty who are talented case writers create more value at Harvard Business School, which has matching idiosyncratic resources (e.g., Harvard Business Publishing), than at other universities. We propose that employee–brand complementarity—the value created from the fit between brand and employee differentiation—is particularly important for horizontally differentiated brands.
The ability of firms and employees to match optimally is challenged by labor market frictions or search costs due to imperfect information. On the firm side, imperfect information makes it costly to identify high-quality matches. Firms can reduce frictions by investing in stronger hiring capabilities (Jiang et al. 2012) or building well-known and respected brands, which would be more salient and attractive to applicants during the job search process (Collins and Han 2004). Conversely, imperfect information makes it costly for applicants to signal that they are high-quality, but they can do so, for example, by obtaining a degree from a respected university (Spence 1973) or working for a prestigious brand (Bidwell et al. 2015). All else equal, the higher the frictions faced by the firm (employee), the lower the firm's (employee's) bargaining power.
Predictions
Figure 1 depicts our framework. We begin by theorizing that pay will be lower (higher) at vertically (horizontally) differentiated brands. We further theorize that the effect of horizontal brand differentiation on pay will be influenced by the degree to which prospective employees are horizontally differentiated in a way that fits the brand. We then consider how firm strategies and investments across the employee–brand matching process (labor availability, identification, attraction, and development) shift firm bargaining power and moderate the brand–pay relationship. Given these effects, we predict that to the extent vertical (horizontal) brand differentiation results in lower (higher) pay, profits will decrease (increase) as a result of the negative (positive) mediating effects of employee productivity and retention.
How Vertical and Horizontal Brand Differentiation Impact Pay and Profits Through Employee–Brand Matching.
The Effect of Vertical Brand Differentiation on Pay
Previous literature on pay has, implicitly or explicitly, described employee-based brand equity in terms of vertical brand differentiation, such as brand status or prestige (Bidwell et al. 2015; DelVecchio et al. 2007; Tavassoli, Sorescu, and Chandy 2014; Yang, Shi, and Goldfarb 2009). For example, Cable and Turban (2003) asked students to rate hypothetical employers on “good public image” and “very familiar” to predict minimum salary requirements. DelVecchio et al. (2007) similarly observed lower salary requirements for hypothetical jobs at firms with more expensive brands that have higher awareness, market share, and perceived quality.
Vertical brand differentiation enhances firm bargaining power to offer lower pay for several reasons. First, a key nonpecuniary benefit affecting firm bargaining power is brand-knowledge transfer from being associated with a firm's brand (Highhouse, Thornbury, and Little 2007; Tavassoli, Sorescu, and Chandy 2014; Yang, Shi, and Goldfarb 2009). Core to this view is the premise that a person's underlying quality is a difficult-to-observe characteristic that is socially constructed (Lynn, Podolny, and Tao 2009). For example, the perceived quality of scientists is based not only on scholarly merit but also on the prestige of their academic affiliations (Hargens and Hagstrom 1982). Hedonic wage theory suggests that this brand-knowledge transfer constitutes psychic wages that substitute for pay (Lucas 1977).
Second, brand-knowledge transfer benefits employees in the form of résumé power, which can also serve as a substitute for pay. Résumé power from a vertically differentiated brand serves as a signal of general human capital that enhances employees’ value in labor markets (Tavassoli, Sorescu, and Chandy 2014). Because firms are uncertain about the quality of an applicant, they consider signals such as the vertical differentiation of an applicant's degree-granting institution (Spence 1973) and former employers (Bidwell et al. 2015) to assess quality.
Third, vertical brand differentiation improves firm bargaining power by reducing search costs associated with labor market frictions. In particular, research has shown that the brand awareness associated with quality perceptions (Bronnenberg, Dubé, and Moorthy 2019) increases the number of job applicants at well-known and respected brands, thereby reducing firm search costs (Collins 2007; Collins and Han 2004). For all these reasons, we expect that the higher a firm's level of vertical brand differentiation, the greater its bargaining power and ability to attract the same quality of employee for less pay.
The arguments in support of H1 do not leverage any link between vertically differentiated employees and brands. This is because both positive assortative matching (Mackey, Molloy, and Morris 2014) and negative assortative matching (Yang, Shi, and Goldfarb 2009) are possible under different conditions (Eeckhout 2018). Positive (negative) assortative matching refers to employees and brands being more (less) likely to match when they share the same characteristics. We offer a detailed discussion of each type of matching in Web Appendix A and simply note here that we do not expect vertical employee differentiation to moderate the effect of vertical brand differentiation on pay in H1. In contrast, we discuss next how positive assortative matching should impact the relationship between horizontally differentiated employees and brands.
The Effect of Horizontal Brand Differentiation on Pay
Whereas previous research has linked vertical brand differentiation to lower pay, the employee-based brand equity literature has been entirely silent on horizontal brand differentiation. We address this shortcoming and predict that horizontal brand differentiation will have opposite effects on pay for the following reasons.
First, horizontal differentiation is, by definition, positively assortative. Horizontally differentiated brands should seek to match with the same type of horizontally differentiated employees because these employees offer brand-relevant complementarities through their specific human capital (Gelb and Rangarajan 2014; Sirianni et al. 2013). These complementarities create greater economic value for the firm, which is shared in the form of higher pay (Becker 1964).
Second, a firm's demand for specific human capital that fits the requirements of its horizontal brand differentiation creates a restrictive matching condition in the form of a smaller available labor pool (Mackey, Molloy, and Morris 2014). As with any scarce talent, this increases the bargaining power of employees who are brand-relevant matches, resulting in horizontally differentiated firms paying higher wages for a match (Campbell, Coff, and Kryscynski 2012; Mackey, Molloy, and Morris 2014).
Third, the same way horizontally differentiated brands appeal to only a subset of consumers (Datta, Ailawadi, and Van Heerde 2017), they should provide psychic wages in the form of self-expression benefits to only a minority of employees (Highhouse, Thornbury, and Little 2007). Brand-knowledge transfer from such brands may even be a disincentive for individuals who do not identify with the unique brand associations. For example, employees who want to express a feminine self-identity may receive negative utility from the brand-knowledge transfer of Wildfang's more masculine/androgynous image. This disincentive restricts the number of potential employee matches attracted to the firm.
Fourth, it follows that horizontal brand differentiation confers limited résumé power because it signals specific human capital that is not transferable to most other firms. Working for Southwest Airlines signals that an employee has a quirky personality, which may make this employee more attractive to a brand like Benefit Cosmetics (with the tagline “Laughter is the best cosmetic”) but not to brands that do not value this quality. For all these reasons, we predict:
The positive assortative matching underlying this main effect further suggests an interaction such that the effect of horizontal brand differentiation on pay should be highest for employees matched on (brand-relevant) horizontal differentiation. We therefore predict:
Employee–Brand Matching Process Moderators
The matching process between employees and brands is shaped by various factors—the firm's demand for labor (labor availability), its approach to creating a consideration set of candidates (labor identification), its ability to attract candidates (labor attraction), and its investment into match quality (labor development). We predict that firm strategies and investments in these activities will shift firm bargaining power and impact the brand–pay relationship. We offer a broad discussion of these moderating factors and test their effects using an array of specific measures.
Labor availability
The employee–brand matching process relies on the supply of talent that meets the firm's demand. All else equal, firms have lower bargaining power when they require human capital that is in short supply relative to firm demand (Mackey, Molloy, and Morris 2014). This dynamic underlies our prediction in H2b. Firms also face labor availability constraints as the number of firms competing for similar talent increases. These types of labor-supply requirements should weaken the firm's bargaining power and decrease (increase) the negative (positive) effect of vertical (horizontal) brand differentiation on pay.
Labor identification
Firms need to identify qualified applicants from the pool of available labor to form a consideration set from which to hire. Firms may rely on different strategies and investments that either enhance or limit their bargaining power. One approach is to build stronger capabilities in people practices, including hiring procedures and compensation negotiations (e.g., Jiang et al. 2012). Such capabilities reduce labor-market frictions and increase the firm's bargaining power (Schmidt and Hunter 1998), thus increasing (decreasing) the negative (positive) effect of vertical (horizontal) brand differentiation on pay.
Firms can also strategically rely on employee referrals to identify and screen candidates. Such employee referrals reduce uncertainty about applicant quality by virtue of the private information an internal referrer has about difficult-to-observe candidate characteristics such as person–organization fit (Eeckhout 2018; Montgomery 1991). Further, referrals are credible signals of quality (Spence 1973) because current employees put their reputations on the line to endorse applicants. This higher certainty and certification increase a candidate's bargaining power. Therefore, the negative (positive) effect of vertical (horizontal) brand differentiation on pay is decreased (increased) by the degree to which firms rely on employee referrals in hiring.
Labor attraction
Once identified, firms compete to attract applicants through pecuniary and nonpecuniary benefits. Our theory has thus far focused on résumé power and psychic wages as the key source of nonpecuniary benefits. However, companies also offer other benefits associated with a positive work environment or with an employee's health and well-being. Such benefits increase firm bargaining power, which should increase (decrease) the negative (positive) effect of vertical (horizontal) brand differentiation on pay.
Labor development
Firms invest in match quality by developing labor through various means, such as on-the-job training. Becker’s (1964) seminal theory of human capital is centered on the idea that there are two types of training that improve human capital: general and specific.
We posit that training that improves vertical brand differentiation can be considered general training because it relates to quality dimensions universally valued by firms and consumers. For example, any training that enhances the job skills of a physician or a customer service agent will be valued by all firms because such skills are valued by all consumers. Given this, general training increases the value of the employee not only at the focal firm but also to many other firms. If so, it is not efficient for firms to fully absorb the costs of such general training, and firms should pass on some of the training costs in the form of lower wages (Becker 1964). Employees, in turn, should absorb these costs in exchange for an increase in their vertical employee differentiation. As a result, general training that relates to vertical brand differentiation should strengthen the negative effect of vertical brand differentiation on pay.
In contrast, training to maintain or improve horizontal brand differentiation is nontransferable to other firms; thus, employees need to be incentivized to invest time and effort to acquire this type of specific training (Becker 1964). Given this, the resulting productivity gains from specific training tend to be shared with employees in higher pay. Specific training also increases the employee's replacement costs at the focal firm (Campbell, Coff, and Kryscynski 2012). For both reasons, higher levels of specific training that relate to horizontal brand differentiation should strengthen the positive effect of horizontal brand differentiation on pay.
Having considered the effect of employee–brand matching on pay, as well as process moderators, we next consider the downstream effect on firm profits. To simplify the presentation, our predictions focus only on the mediating effect of pay and associated intermediate employee productivity and retention outcomes on the brand–profit relationship. However, we expect the aforementioned moderators on the brand–pay relationship to flow through to profits, which we examine using tests of moderated mediation.
The Impact of Brand-Based Pay on Profits
Previous research has documented the negative effect of brand equity on pay and suggested that if brand equity is leveraged to lower pay, this should translate into higher profits (Tavassoli, Sorescu, and Chandy 2014). We challenge this view. Specifically, to the extent that vertical brand differentiation is leveraged to lower pay, we expect it to diminish profits due to losses in employee productivity (i.e., rate of output per employee) and retention (i.e., proportion of employees who voluntarily chose to stay). In contrast, when managers at firms high on horizontal brand differentiation pay more, profits increase via the same mediating employee behaviors.
This view is supported by the efficiency wage literature, which points to the motivational effects of pay on productivity. As Akerlof and Yellen (1990, p. 258) note, “the motivation for the fair wage-effort hypothesis is a simple observation concerning human behavior: when people do not get what they deserve, they try to get even.” Employees may not anticipate that accepting lower pay at a vertically differentiated brand will become demotivating despite the psychic wages or résumé power they receive in return (Deci, Koestner, and Ryan 1999).
Conversely, higher pay can be motivating. Efficiency wage research has shown higher pay to result in employees exerting extra effort and reducing shirking due to higher morale, reciprocity, and perceived fairness (Akerlof and Yellen 1986; Weiss 2014), thereby driving up productivity and profits. For example, Henry Ford's decision to offer workers $5 a day in 1914 (nearly twice the going rate) led to higher profits via productivity gains (Raff and Summers 1987). Government workers have also been found to increase effort levels by 1% per .24% increase in pay (Taylor and Taylor 2011).
Brand-based pay should also affect employee mobility. All things equal, the lower the pay, the more attractive are outside job opportunities. This should result in lower retention when the firm has leveraged its vertical brand differentiation to lower pay because employees can expect (and may seek) higher pay elsewhere. The opposite dynamic should unfold when horizontal brand differentiation translates into higher pay—these employees should, on average, expect a pay cut from switching jobs to join firms that may not value their specific human capital (Campbell, Coff, and Kryscynski 2012). This creates a bilateral monopoly dynamic (mutual hostage situation) because higher specific human capital simultaneously increases and limits employee bargaining power. As a result, employees will be paid more, and the incumbent employer enjoys higher profits because employees cannot entirely appropriate the economic rents they produce (Campbell, Coff, and Kryscynski 2012; Mackey, Molloy, and Morris 2014). Finally, a higher-than-market-rate pay has also been shown to increase job satisfaction, which decreases the attractiveness of outside offers and increases retention (Galizzi and Lang 1998).
For all these reasons, lower (higher) pay due to vertical (horizontal) brand differentiation should result in lower (higher) levels of employee productivity and retention and thereby attenuate (enhance) profits. Research on efficiency wage models indicates that managers do not anticipate the full impact of pay on productivity or retention (Akerlof and Yellen 1986; Weiss 2014). This is likely because, unlike discrete compensation costs, these costs occur over time and manifest in an array of absenteeism, productivity losses, separation, and replacement costs that are difficult to measure accurately (Cascio 1982). We therefore predict:
Archival Data Strategy
Overview of Empirical Strategy
We use a multimethod approach to test our framework. We rely on archival data to test the effect of brand differentiation on pay (H1, H2a), the effect of brand differentiation on profits as mediated by pay and employee behaviors (H3, H4), and how firm strategies and investments across the employee–brand matching process shift firm bargaining power to influence the brand–pay relationship. We use experiments to complement this analysis in three ways. First, we test the validity of our archival measures of vertical and horizontal brand differentiation in a study with consumers. Second, we test the effect of employee–brand matching on pay (H2b) in two studies with human resource (HR) managers. Third, we test our assumptions about manager (employee) myopia regarding the effects of pay on productivity and retention in two studies using HR managers (students).
Data Description and Sample
Testing our predictions requires firm-level data on brand differentiation, pay, productivity, retention, and profits. We use Brand Asset Valuator (BAV) data from Young & Rubicam to measure brand differentiation. We obtain data on pay, retention, and various other controls from applications to Fortune's 100 Best Companies to Work For list (hereinafter Fortune “best places to work” list). Collected by The Great Place to Work Institute (GPWI), these data rely on two surveys. The Trust Index© examines management practices and company climate and is completed by a random sample of 200 employees from each submitting company. The Culture Audit© examines questions about average pay, retention, and benefits and is completed by HR professionals. We use additional databases to construct moderators and control variables as detailed next.
Our sample is constructed by intersecting data obtained from BAV, GPWI, and Compustat. Specifically, the intersection of BAV data from 930 firms, GPWI data (2006–2017) from the 628 publicly traded firms that applied, and Compustat data resulted in 526 observations from 183 public firms. This sample constitutes an unbalanced panel, with 87 firms applying only once to the Fortune “best places to work” list and only five firms applying in all 12 years. The average firm in our sample is relatively large, with over 16,000 employees, which is not surprising given the Fortune application process requires resources that are most likely to be found in larger firms.
Dependent Variables
Our ultimate dependent variable is profits, which is measured using the logarithm of earnings before interest, tax, depreciation, and amortization (EBITDA) from Compustat. This type of accounting-based metric of firm performance is a more appropriate outcome of the effect of brand differentiation on pay and employee behaviors than forward-looking metrics of firm value (e.g., Tobin's q, which we nevertheless include in a robustness test). This is because the effect of pay, retention, and productivity primarily materializes in the annual performance of the firm. We therefore use profits at time t + 1 and all independent variables at time t to allow for the effects of pay, productivity, and retention to be recognized in firm performance.
Independent Variables
Overview of measures of brand differentiation
We rely on BAV data to construct our measures of horizontal and vertical brand differentiation. These data are based on the perceptions of a representative sample of U.S. adults and include measures of brand knowledge that allow for comparisons across products and markets. Horizontal brand differentiation, defined as the holistic perception of a brand's uniqueness relative to other brands, is measured as the average of the (standardized) percent agreement with “unique,” “different,” “distinctive,” and “dynamic” brand associations. Vertical brand differentiation, defined as the holistic perception of a brand's superior quality relative to other brands, is captured by six measures: the percent agreement with “high quality,” “leader,” and “reliable” brand associations and ratings of “regard” (on a seven-point scale anchored by “Extremely low regard” and “Extremely high regard”); ratings of “familiarity” (on a seven-point scale anchored by “Never heard of” and “Extremely familiar”); and ratings of “relevance” (on a seven-point scale anchored by “Not at all relevant” and “Extremely relevant”). 2 Vertical brand differentiation is the average of these six (standardized) items.
Support from literature for our brand differentiation measures
We offer evidence from the consumer behavior and brand equity literatures to support our measures. The consumer behavior literature examining horizontal and vertical differentiation has relied on similar types of definitions, measures, and manipulations (see Web Appendix B and Table WB). Consumer research has measured horizontal differentiation as “atypical,” “unrepresentative,” and “dissimilar” (Dommer, Swaminathan, and Ahluwalia 2013, p. 661) and found it to be associated with Google Trends search terms such as “alternative, anti-establishment, Bohemian, counterculture, creative, distinct, geek, Indie, unconventional, unorthodox” (Ordabayeva and Fernandes 2018, p. 245). Research has also manipulated horizontal differentiation through “just different” and “unique” (Ordabayeva and Fernandes 2018, p. 233). Consumer research has also associated “dynamic” with horizontal brand differentiation. For example, Ordabayeva and Fernandes (2018, p. 245) measure “hipster” and “creative” in Google search terms and manipulate a brand description as “edgy and irreverent, hip.”
The literature also supports the idea that “high quality,” “leader,” “regard,” and “reliable” reflect vertical brand differentiation. Studies measure ratings of “status, wealth, power, and prestige” (Dommer, Swaminathan, and Ahluwalia 2013, p. 661), whether products are “better” (Spiller and Belogolova 2017, p. 974), and “better/best, elegant, elite, luxury, money, prestige, rich, and success” from Google search terms (Ordabayeva and Fernandes 2018, p. 245). Research also manipulated “just better” (Ordabayeva and Fernandes 2018, p. 233) and “objectively better” (Spiller and Belogolova 2017, p. 974).
Previous literature also included “familiarity” as an important component of vertical brand differentiation. Cable and Turban (2003, p. 2256) find that ratings of company familiarity (e.g., “I am very familiar with this firm”) are a strong predictor of perceptions of its reputation—an indicator of vertical brand differentiation. DelVecchio et al. (2007, p. 154) measure “brand awareness” and “perceived quality” to reflect brand strength—another indicator of vertical brand differentiation. We believe that familiarity contributes to vertical brand differentiation for two reasons. First, research has documented a positive effect of brand awareness and advertising expenditures on perceptions of quality (e.g., Bronnenberg, Dubé, and Moorthy 2019), even after accounting for actual quality (Moorthy and Zhao 2000). Second, for consumers to use vertically differentiated brands to signal their own quality (Ordabayeva and Fernandes 2018), the brand must be recognized by others. 3 In contrast, horizontal brand differentiation can result from both low and high familiarity: low, because consumers tend to perceive unfamiliar brands as atypical of a category (Bijmolt et al. 1998) and high, when consumers become familiar with a brand's unique attributes (Murphy and Wright 1984).
Turning to relevance, which reflects how important a brand is across individuals in the marketplace, the literature points to its role in vertical brand differentiation for the following reasons. First, relevance is a motivational variable (Celsi and Olson 1988; Zaichkowsky 1985) and more involved consumers have been found to hold stronger quality beliefs (Steenkamp 1990). Second, consumers can also better signal their superiority in product domains that are widely relevant because it is easier to make attribute-based comparisons across quality levels (Liu, McFerran, and Haws 2020). For example, the practice of “keeping up with the Joneses” refers to people wanting to own the same brands as their peers in order to keep pace with them. Third, highly relevant brands tend to have larger market shares, and market share has been shown to be a quality signal (Bhattacharya, Morgan, and Rego 2022). DelVecchio et al. (2007) also include “market share” as a measure of brand strength. In contrast, horizontal brand differentiation should be lower when it is relevant to many consumers. Consistent with this idea, individuals with a higher need for uniqueness have been shown to differentiate themselves by rejecting choices deemed relevant by others (Berger and Heath 2007).
In addition to the consumer behavior literature, the brand equity literature relying on BAV data suggests that the BAV pillars of esteem (composed of high quality, leader, regard, and reliable), relevance, and familiarity are related to vertical brand differentiation, whereas energized differentiation (composed of unique, different, distinctive, dynamic, and innovative) is related to horizontal brand differentiation, albeit without invoking these construct labels (see Web Appendix C for details). Datta, Ailawadi, and Van Heerde (2017) show that energized differentiation correlates negatively with sales-based brand equity and market share because it “does not necessarily appeal to the masses” except in hedonic categories where consumers can “better ascertain and appraise a brand's unique aspects” (p. 13). In contrast, esteem, familiarity, and relevance positively correlate with sales-based brand equity and market share, especially in categories with “social value … because these brands are more likely to be recognized and respected by others” (p. 6). Similarly, Lovett, Peres, and Shachar (2013, 2014) observe that brand esteem, relevance, and familiarity are related to brand visibility and customer perceptions of satisfaction, usage, and lower risk.
Empirical support for our measures of brand differentiation
We offer three types of empirical evidence to support our measurement approach. First, we perform an exploratory and confirmatory factor analysis of our measures. As detailed in Web Appendix D, our results show that our measures are reliable and have discriminant and convergent validity.
Second, we conduct an experiment that establishes the correspondence between the items used in our measures and the constructs of vertical (perceptions of superior product quality) and horizontal (perceptions of uniqueness) brand differentiation (Study 1 in Web Appendix E).
Third, we use external YouGov brand measures and our archival data to establish nomological validity (see Web Appendix F). We find that YouGov measures of brand quality (as an indicator of quality), brand recommendation (as an indicator of regard), brand awareness (as an indicator of familiarity), brand consideration (as an indicator of relevance), and brand health (which is the sum of these metrics and other metrics such as buzz and awareness of firm advertising as an overall indicator of regard and quality) correlate more strongly with our measure of vertical brand differentiation than with our measure of horizontal brand differentiation (Table WF.1). Further, consistent with their more universal appeal, firms high in vertical brand differentiation are larger and have higher market shares. In contrast, employees at firms high in horizontal brand differentiation respond more positively to the Trust Index© measure “I can be myself around here,” which indicates positive assortative matching (see Table WF.2).
Mediator Measures
Pay is obtained from HR professionals in the Culture Audit. Specifically, they are asked to identify “What is the job function or title of the largest number of full-time salaried employees?” followed by “What was the average annual base pay rate for an employee in this position in the past 12 months?” While we were not able to obtain the exact job function for which salaries are reported (most respondents left the field blank), the manner in which these data are collected ensures that the reporting is done for the most representative group of employees for each firm. We also replicate our findings using a measure of pay that includes both salary and bonus. Consistent with Tavassoli, Sorescu, and Chandy (2014), we log-transform these variables.
We use data envelopment analysis (DEA) to create our measure of productivity (e.g., Donthu and Yoo 1998; Kamakura, Ratchford, and Agrawal 1988). This captures the extent to which the firm produces the maximum quantity of outputs for a given level of inputs, where the frontier is determined by the set of firms in the same three-digit Standard Industrial Classification (SIC) code. Following Bucklin (1978) and Doutt (1984), we use the number of employees and assets as inputs, net profit margin as outputs, and the DEA package in Stata to compute each firm's productivity score. We use net profit margin as the performance output to avoid a direct correlation of this output with EBITDA—our ultimate dependent variable. We replicate our findings with a productivity measure based on employee ratings of “People here are willing to give extra to get the job done” from the Trust Index.
We measure retention as the percentage of employees who voluntarily remained with the company in a given year, reflecting 100% minus voluntary turnover reported by HR professionals as part of the Culture Audit. We replicate our findings with a retention measure based on employee ratings of “I want to work here a long time” from the Trust Index.
Measures of Employee–Brand Matching Process Moderators
Our conceptualization of the employee–brand matching process offers a broad discussion of factors that moderate the brand–pay relationship. Next, we describe six specific moderators we expect to impact firm bargaining power (see Web Appendix G for predictions).
Labor availability
We examine two moderators that decrease firm bargaining power due to its demand for specific types of labor. First, demand for technical talent, defined as firm hiring requirements for employees with high levels of specialized analytical, engineering, and scientific skills, is captured by three indicators that constitute a formative measure: the average of the standardized values of R&D intensity (Compustat), the relative number of new product announcements made by each firm (RavenPack), and a dummy that captures whether the firm is listed on Nasdaq (Center for Research in Security Prices), which is focused on new technologies (see Web Appendix H).
Second, demand for frontline employees, defined as firm hiring requirements for employees who work at the boundary of the organization and directly represent the brand to customers, is important in services industries (Eckert et al. 2022; Vomberg, Homburg, and Bornemann 2015). We measure it with a dummy that is 1 for services firms and 0 otherwise.
Labor identification
We examine two moderators that should influence firm bargaining power due to their ability to create a strong consideration set of potential employees. First, human resource management (HRM) sophistication, defined as the degree to which the firm is competent and innovative in people practices (Jiang et al. 2012), should increase firm bargaining power. This is measured as the degree to which the firm has developed four key HR policies: a health and safety policy, a diversity and opportunity policy, a policy against child labor, and a policy supporting the human rights of employees (collected by Eikon Refinitiv as part of a governance assessment). We average the four scores, which range from 0 to 100 (α = .83). Second, reliance on employee referrals, which should decrease bargaining power, is reported by HR professionals in the Culture Audit as the percentage of new hires referred by current employees.
Labor attractiveness
Benefits, which increase firm bargaining power, are measured using the percentage of nine benefits offered by the company (e.g., childcare, tuition, number of days off) (α = .87), as reported in the Culture Audit and summarized in Web Appendix H.
Labor development
Employee training comes from the Culture Audit and is the number of hours of training offered to the largest group of employees.
Control Variables
In addition to industry and time dummies, we include 11 control variables to rule out endogeneity threats due to observable determinants of our dependent variables. 4 We include vertical (horizontal) brand differentiation in models where horizontal (vertical) differentiation is the focal independent variable because brands can vary on both, 5 firm size using the logarithm of the number of employees from the Culture Audit data; the proportion of managers among the respondents to the Trust Index to capture employee heterogeneity, and a dummy for corporate brand because the effect of brand on pay may be stronger for these firms.
We include two controls that reflect the overall positive nature of the workplace: (1) employee engagement based on the average employee response to 12 questions that reflect engagement and satisfaction with the workplace from the Trust Index (summarized in Web Appendix H) and (2) firm inclusion on the Fortune's “best places to work” list (using a dummy variable). We also use the previously described benefits and training as control variables in the pay and employee behavior models because employees likely derive utility from them, which can impact the pay they accept and their productivity and retention.
We include industry concentration measured by the Herfindahl–Hirschman index as the sum of squares of firm market share in an industry defined by three-digit SIC codes (e.g., Luo, Homburg, and Wieseke 2010). We add a dummy that captures whether the firm is in the higher-paying high-tech sector (SIC codes 35, 59, and 73; Mizik and Jacobson 2009) 6 and industry sales growth to the profit equation because higher-growth industries provide more opportunities to earn profits (Lee 2014). Table 1 describes all variables in our models.
Descriptive Statistics.
Notes: All variables are at the firm level unless noted. Correlations higher than .09 are significant at 95% or higher.
Archival Data Estimation Approach
Models
Serial mediation model
Our theory posits that vertical and horizontal brand differentiation have opposite effects on pay, which influences employee productivity and retention to affect profits. This structure reflects a serial mediation model, depicted in Figure 1, which we test using PROCESS Model 81 with 5,000 bootstrapped samples (Hayes 2013). We control for selection bias and endogeneity for observable firm differences in the choice to apply to the Fortune ranking with a Heckman selection model and for the endogeneity of brand differentiation, pay, productivity, and retention with a control function approach (all discussed subsequently).
We do not control for unobserved firm differences using a fixed-effects model for our main mediation models because 87 of the 183 firms in our sample have only one observation. We do, however, add industry and year fixed effects to all models and control for a range of observables. Moreover, as we discuss subsequently, we use an instrument-based correction for endogeneity, which helps account for the effect of unobservables (Wooldridge 2015). We measure all independent variables and mediators at time t and profits at time t + 1 (while controlling for profits at time t). Using this specification, we fail to reject the null hypothesis of no first-order autocorrelation using a Wooldridge test (F(1, 56) = .20, p = .66).
We estimate the following system of four equations:
where i denotes firm, t denotes year, and λ controls for the potential selection bias caused by only including firms that applied to the Fortune “best places to work” list. We explain in the next section how we obtain Resid_VBD and Resid_HBD, which control for the endogeneity of vertical and horizontal brand differentiation; Resid_Pay, which controls for the endogeneity of pay; and Resid_Productivity and Resid_Retention, which control for the endogeneity of productivity and retention. The remaining variables are as previously described.
We take the following steps to increase confidence that our profit results are due to mediation. First, key predictors (brand differentiation), moderators (across the employee–brand matching process), mediators (pay and employee behaviors), and outcomes (profits) represent distinct theoretical domains. Second, the intercorrelations between variables are low (see Table 1). Third, the variables are measured using distinct approaches that leverage data from different sources, reducing common method bias. Fourth, our brand and employee moderators and mediators are temporally separated from profit, which is assessed in t + 1.
Moderated serial mediation model
We examine the impact of six moderators that reflect facets of the employee–brand matching process on the brand differentiation–pay relationship and on the mediating effect of brand differentiation on profits by adding them to Equation 1 and using PROCESS Model 83 with 5,000 bootstrapped samples. The statistical inference drawn from these models refers to whether each moderator has a nonzero weight in the function linking the indirect effect of brand differentiation on profits through employee behaviors (Hayes 2015). This weight, referred to as the index of moderated mediation (IMM), is reported in Table 5, and the estimation of moderated serial mediation models is described in Web Appendix I.
Identification Strategy
It is important to account for potential sources of bias that threaten the identification of our models. First, we need to account for firm choice to apply to the Fortune “best places to work” list. Second, we need to account for the endogeneity of brand differentiation, pay, productivity, and retention. We briefly outline our identification strategy here and offer a complete description of all instruments, explanations of their validity, our models, and results in Web Appendix J.
Accounting for firm choice to apply for inclusion in the Fortune list
Firms do not randomly apply to the Fortune list. Instead, the choice is likely driven by firm and industry characteristics that could introduce selection bias. We control for this potential bias by using a Heckman selection model (see Equation 5 in Web Appendix J) in which we estimate the decision to apply to the ranking (see Table WJ.1) and then use the inverse Mills ratio obtained from this model as a control variable in Equations 1–4.
Accounting for the endogeneity of brand differentiation
The relationships between brand differentiation, pay, employee behaviors, and profits may also be impacted by unobserved variables. For instance, the impact of brand differentiation on profits may depend on market trends that may lead to some brand characteristics being preferred to others. To account for this potential source of endogeneity, as shown in Equations 6 and 7 in Web Appendix J, we use a control function approach that models each brand differentiation dimension as a function of an instrument and a set of control variables (Petrin and Train 2010) (see Table WJ.2). We then extract residuals from these models and add them to Equations 1–4.
Accounting for the endogeneity of pay, productivity, and retention
Given that firms choose the pay they offer employees, pay is endogenous. Our productivity and retention equations may also suffer from endogeneity resulting from unobservables that influence both pay and employee behaviors, such as the threat of a recession or other macroeconomic factors. To account for these possibilities, we use a control function approach that models pay, productivity, and retention as a function of instruments and a set of control variables (see detailed explanations and Equations 8, 9, and 10 in Web Appendix J and Tables WJ.3 and WJ.4). We add residuals for pay to Equations 2–4 and for productivity and retention to Equation 4.
Archival Data Results
How Brand Differentiation Affects Pay
In terms of model-free evidence, we find that vertical brand differentiation is negatively correlated with pay (ρ = −.15, p < .001) and horizontal brand differentiation is positively related to pay (ρ = .16, p < .001). In terms of formal model testing and in support of H1, and as shown in Table 2, results show that vertical brand differentiation has a negative effect on pay (β1 = −.107, p < .01). In contrast, and in support of H2a, horizontal brand differentiation has a positive effect on pay (β2 = .091, p < .001).
The Effect of Brand Differentiation on Pay.
*p < .05. **p < .01. ***p < .001.
Notes: All variables are at the firm level unless noted. The model includes industry and year dummies, which are not reported for parsimony.
How Brand Differentiation Affects Profits as Mediated by Pay and Productivity
We now consider the effects of brand differentiation on profits through the mediating effects of pay and productivity. Using PROCESS Model 81, we test the pathways in Figure 1 for brand differentiation through to profits. Table 3, Parts 1a and 2a, report all direct pathways, and Parts 1b and 2b report the serial mediation effects in the form of indirect effects from brand differentiation to profits through pay and productivity.
Brand Differentiation → Pay → Productivity → Profits Mediation Results.
Notes: LL = lower limit of the confidence interval; UL = upper limit of the confidence interval; VBD = vertical brand differentiation; HBD = horizontal brand differentiation.
In support of H3a, vertical brand differentiation has a negative indirect effect on profits mediated by pay and productivity (vertical brand differentiation → pay → productivity → profits indirect effect = −.0095, 95% CI: [−.0258, −.0015]). Importantly, given the direct effect of vertical brand differentiation on profits is positive (see Table 3) and the indirect effect through pay and the employee behaviors is negative, we have a case of competitive mediation (Zhao, Lynch, and Chen 2010). This means that when vertically differentiated brands offer lower pay, the positive effect of vertical brand differentiation on profits is weakened due to a reduction in productivity.
In support of H3b, horizontal brand differentiation has a positive indirect effect on profits mediated by pay and productivity (horizontal brand differentiation → pay → productivity → profits indirect effect = .0081, 95% CI: [.0021, .0180]). Offering higher pay enhances the positive effect of horizontal brand differentiation on profits due to an increase in employee productivity.
How Brand Differentiation Affects Profits as Mediated by Pay and Retention
Using the same approach, we now consider the effects of brand differentiation on profits through the mediating effects of pay and retention. Table 4, Parts 1a and 2a, report all direct pathways, and Parts 1b and 2b report indirect effects. In support of H4a, vertical brand differentiation has a negative indirect effect on profits as mediated by pay and retention (vertical brand differentiation → pay → retention → profits indirect effect = −.0120, 95% CI: [−.0285, −.0021]). As with productivity, this competitive mediation indicates that the positive effect of vertical brand differentiation on profits is weakened due to pay-induced losses in retention.
In support of H4b, horizontal brand differentiation has a positive indirect effect on profits as mediated by pay and retention (horizontal brand differentiation → pay → retention → profits indirect effect = .0103, 95% CI: [.0021, .0237]). Offering higher pay enhances the positive effect of horizontal brand differentiation on profits due to an increase in employee retention.
Brand Differentiation → Pay → Retention → Profits Mediation Results.
Notes: LL = lower limit of the confidence interval; UL = upper limit of the confidence interval; VBD = vertical brand differentiation; HBD = horizontal brand differentiation.
Following Pieters (2017), we reverse the order between pay and the two employee behavior mediators (productivity and retention). As expected, we find that the indirect effect is not significant in any of these four serial mediation models (see Tables 3 and 4).
Employee–Brand Matching Moderators and Moderated Mediation to Firm Profits
We report our findings in two steps. First, we report how moderators associated with the employee–brand matching process influence the brand–pay link (see Web Appendix G for formal predictions). Second, we present evidence regarding whether this moderation, in turn, affects the mediating effect of pay through employee productivity and retention to profits (see Table 5). Specifically, for each moderator, we follow Hayes (2015) and compute the IMM to test whether the moderator influences the paths from brand to profits for each combination of brand differentiation (vertical and horizontal), pay, and employee behavior (productivity and retention). Web Appendix K presents the conditional indirect effects from these models and details how mediation varies across different levels of each moderator.
Moderated Mediation Effects: How the Employee–Brand Matching Moderators Influence the Effect of Brand Differentiation on Pay Through Employee Behaviors to Profits.
Notes: n.s. = not significant.
Labor availability
We find that demand for technical talent positively moderates the effect of vertical brand differentiation on pay (
We find that demand for frontline employees has a positive moderating effect on the effect of horizontal brand differentiation on pay (
Labor identification
As firms improve their ability to identify the right applicants, this should increase firm bargaining power to keep pay low. In support of this view, we find that the moderating effect of HRM sophistication on the effect of vertical (
In contrast, we find that only the moderating effect of employee referrals on the effect of horizontal brand differentiation on pay is positive (
Labor attraction
We predicted that the effect of brand differentiation on pay should depend on firm benefit levels. In support of this view, we find that benefits magnify the negative effect of vertical brand differentiation on pay (
Labor development
We predicted that training that serves vertical (horizontal) brand differentiation should increase (decrease) firm bargaining power. The interaction of training with each brand dimension allows us to identify these opposing effects. Training has a negative moderating effect on the effect of vertical brand differentiation on pay (
Additional Analyses
Ruling Out Alternative Mechanisms
Our theory focuses on bargaining power dynamics between firms and employees as the underlying mechanism for the brand–pay relationship. We test this directly in our experiments, which follow, and indirectly in our examination of the effect of a set of moderators associated with the employee–brand matching process in our archival data. Web Appendix L presents tests ruling out four alternative explanations—benefits and training, firm resources, and employee age as explanations for the negative effect of vertical brand differentiation on pay, and employee diversity as an explanation for the positive effect of horizontal brand differentiation on pay.
Robustness Checks
We replicate our results using 11 different tests, including different measures of brand differentiation, pay, productivity, retention, and profits (see Web Appendix M).
Sustainability of Firms Using a Vertical Brand Differentiation Low Pay Strategy
Our results indicate that when managers leverage vertical brand differentiation to lower pay, it negatively affects profits due to mediating effects of lower employee productivity and retention. To investigate the sustainability of this strategy over the long run, we show in a series of alternative models that this lower pay does not have a negative impact on vertical brand differentiation or sales in future time periods. Instead, perceptions of vertical brand differentiation appear to be sticky over time (see Web Appendix N for results).
Experimental Results
Our archival data lack direct measures of employee differentiation. Hence, we cannot test H2b. We therefore conduct two experiments that manipulate employee differentiation to test our hypotheses. Study 2 examines employee–brand matching to test our assumption of no assortative matching for vertical differentiation and positive assortative matching for horizontal differentiation in H2b. Study 3 examines the interaction of horizontal brand and employee differentiation to provide an additional test of H2b. We also conduct Studies 4 and 5 to examine whether managers and job candidates are myopic about the demotivating effects of lower pay.
Study 2: Employee–Brand Matching Experiment
Study 2 examines whether employee–brand matching is assortative. It relies on a sample of 204 HR managers from chapters of the Society for Human Resource Managers (SHRM) (Mage = 36.15 years; 149 female, 55 male, 0 other; 7.2 years of HR experience) and employs a hypothetical hiring scenario using a 2 (high vertical or horizontal employee differentiation) × 2 (high vertical or horizontal brand differentiation) design. A repeated-measures general linear model analysis reveals a significant brand-by-employee interaction. Examining the interaction, we find that HR managers at firms high in horizontal brand differentiation are willing to pay more for horizontal employee matches than for highly vertically differentiated employees. In contrast, HR managers at firms with high vertical brand differentiation do not offer highly vertically or horizontally differentiated employees differential pay. These results, which are discussed in detail in Web Appendix P, support the idea that vertical matching is nonassortative, whereas horizontal matching is positively assortative (H2b) and limits firm bargaining power.
Study 2 also tests our assumptions regarding résumé power, psychic wages, and the relative scarcity of horizontal employee–brand matches. We find that vertical brand differentiation confers more résumé power, but not social status, than high horizontal brand differentiation, while horizontal brand differentiation provides self-expressive benefits. Importantly, this study also shows that HR managers perceive employees who provide a high horizontal brand fit to be scarcer than employees who are highly vertically differentiated.
Study 3: Horizontal Matching Experiment
While Study 2 examines H2b by examining vertical versus horizontal matching, Study 3 tests H2b using different levels of horizontal brand and employee differentiation. It utilizes a hypothetical hiring scenario in a 2 (low or high horizontal brand differentiation) × 2 (low or high horizontal employee differentiation) design. The sample consisted of 118 HR managers recruited via SHRM (Mage = 47.4; 104 female, 14 male, 0 other; 14.6 years of HR experience). A repeated-measures ANOVA indicates a significant brand-by-employee interaction. Consistent with H2b, we find that HR managers at firms with high (low) horizontal brand differentiation are (are not) willing to pay more for horizontally matched employees because of lower perceived firm bargaining power (see Web Appendix Q for details). Study 3 thereby demonstrates positive assortative matching across different levels of horizontal brand and employee differentiation.
Study 4: Are HR Managers Myopic About the Effects of Pay?
Study 4 tests the assumption that HR managers do not fully anticipate the behavioral consequences of paying less at firms with vertically differentiated brands. Ninety-five HR managers (Mage = 48.9 years; 89 female, 6 male, 0 other; 15.1 years of HR experience) recruited via SHRM participated in a 2 (high or average vertical brand differentiation) × 2 (10% below- or industry-average pay) study design using hypothetical scenarios. A repeated-measures GLM analysis reveals no main effect of pay on productivity and a significant main effect of vertical brand differentiation with managers expecting employees to work harder at the high vertical brand firm. Importantly, we do not observe a significant pay-by-vertical brand differentiation interaction on productivity expectations, indicating that managers are myopic about the behavioral consequences of lower pay at vertically differentiated brands. A parallel analysis for the retention measure reveals similar main effects. Managers are, however, sensitive to the effects of pay on retention, but only marginally so (see Web Appendix R).
Study 5: Are Job Candidates Myopic About the Effects of Pay?
Study 5 examines job candidates’ willingness to accept lower pay at firms with vertically differentiated brands and whether they anticipate they will be less productive and stay for a shorter period if they do so. A sample of 129 students (Mage = 20.7 years; 73 female, 54 male, 2 other) participated in a 2 (high or average vertical brand differentiation)× 2 (10% below- or industry-average pay) study design. In a hypothetical scenario, we find that job candidates are more likely to accept a job at industry-average (vs. below-average) pay levels and at a highly (vs. average) vertically differentiated brand. Further, a significant brand-by-pay interaction shows a smaller decrease in job acceptance due to lower pay at firms with high (vs. average) levels of vertical brand differentiation. Furthermore, job candidates do not expect pay levels to affect their productivity or job tenure at either level of vertical brand differentiation, implying that they are myopic about the effects of pay on their future behavior (see Web Appendix S).
Discussion and Implications
Theoretical Implications
Brand differentiation and employee-based brand equity
The brand equity literature has only examined the effects of vertical brand differentiation on pay. This is surprising given that the consumer behavior and branding literatures have shown vertical and horizontal brand differentiation to have opposite effects. Extending this view, we offer a theoretical framework that predicts effects for both vertical and horizontal brand differentiation on pay. This more complete understanding of employee-based brand equity has important implications for the study of branding, including how firms can leverage brands in attracting and retaining employees.
Employee–brand matching
We stake new territory in the literature by leveraging the fact that employees are also differentiated to consider how employee–brand matching influences pay. Our theory and experimental findings indicate that horizontally differentiated brands will pay more for matches with horizontally differentiated employees who offer valuable brand complementarities. Matching vertically differentiated brands and employees is not expected to be uniformly negatively or positively assortative; it depends on employee–firm complementarities (as well how management trades off hiring more vs. better workers; Eeckhout and Kircher 2018).
The employee–brand matching process
We extend the human capital literature by conceptualizing an employee–brand matching process and examining six specific strategies and investments a firm can use to influence its bargaining power across the four different stages of this process. Our findings identify conditions when the brand–pay relationship shifts depending on these firm actions. Highlighting one such effect, Becker (1964) posits that training related to vertical (horizontal) brand differentiation should reflect firm investments in developing general (specific) human capital and therefore increase (decrease) firm bargaining power in setting pay. Consistent with this prediction, we find that training related to vertical (horizontal) brand differentiation magnifies the negative (positive) impact of brand differentiation on pay.
The effect on employee-based brand equity on firm performance
Previous research equated higher profits with the lower pay associated with brand equity (Tavassoli, Sorescu, and Chandy 2014). Our findings challenge this view. Specifically, we find that brand-induced lower pay weakens employee productivity and retention, which lowers profits. Thus, leveraging vertical brand differentiation to lower pay represents a false economy. At the same time, the higher pay due to horizontal brand differentiation results in a net-positive effect on profits via improved employee productivity and retention. It is therefore important to consider the full effect of leveraging brand equity on employee behaviors to evaluate its effect on firm performance.
A new perspective on BAV pillars
As reviewed in Web Appendices B and C, previous consumer and brand research has found brand dimensions to diverge in their effects on consumer and investor behavior, most often finding that the BAVs pillar of energized differentiation (closely related to horizontal brand differentiation) behaves differently from esteem, relevance, and familiarity (associated with vertical brand differentiation). Given our findings and validation work—including replications using the BAV pillars (Web Appendix M, Replication 7)—we believe there is an opportunity to reconsider research on the consequences of brand knowledge using the theoretical lens of vertical and horizontal brand differentiation.
Practical Implications
Avoid leveraging vertical brand differentiation to reduce pay
Our results suggest that even if vertical brand differentiation can be leveraged to reduce pay, managers should refrain from doing so. Instead, managers should consider other ways to leverage their vertical brand differentiation to help their firms, including attracting a larger pool of applicants or highlighting résumé power to convert offers.
Incorporate horizontal brand differentiation into pay benchmarks
HR departments rely on pay benchmarks. Our findings suggest they should consider brand differentiation in setting those benchmarks. On one hand, managers should not be deterred from investing in pay if their firms enjoy high horizontal brand differentiation, since this can translate into higher profits through positive employee behaviors. On the other hand, managers at firms with low horizontal brand differentiation should withstand pressures to increase pay to the level of more horizontally differentiated competitors because they lack the brand complementarities that justify higher pay.
Manage the marketing–HR interface across the employee–brand matching process
Our findings show how brand differentiation, which is typically under the purview of marketing, has a profound effect on HR outcomes. Given that cross-functional cooperation between marketing and HR to build brands is reported to be only moderate (The CMO Survey 2023), marketing managers have an opportunity to work more closely with their HR counterparts to build and leverage the brand across the employee–brand matching process. For example, marketing investments into vertical brand differentiation that increase awareness should consider the returns from increasing the size of the pool of job applicants. Marketing skills should also be used to promote vertical differentiation benefits in the form of psychic wages and résumé power to prospective employees in order to convert job offers. Finally, marketing should be involved in defining the key qualities of employee–brand matches and in designing training to support horizontal brand differentiation.
Future Research Directions
Do other firm characteristics moderate the effect of brand differentiation on pay?
We examine a range of firm moderators impacting firm bargaining power in the employee–brand matching process. Future research might consider other factors that occur in the matching process—for example, the impact of firm location, which may influence access to labor, or stronger coordination between marketing and HR, which should improve firm bargaining power.
Do employee characteristics moderate the effect of brand differentiation on pay?
We focus on employee differentiation as reflected in general and specific human capital. Future research could examine whether the effect of brand differentiation on pay varies by employee type. For example, women and minorities tend to earn lower pay (have less bargaining power). It would be interesting to know whether these populations are more likely to take a pay cut to work at vertically differentiated brands and whether this translates into an overall positive payoff for their careers due to the associated résumé power. Further, future research could examine if Ordabayeva and Fernandes’s (2018) finding that vertical (horizontal) brand differentiation has more utility to consumers with a conservative (liberal) ideology also applies to employees.
What other factors might mediate the brand → profit relationship?
Although we find that pay and employee behaviors mediate the relationship between brand and profits, given the direct effect of vertical brand differentiation on profits, other mediators are also operating (Zhao, Lynch, and Chen 2010). Future research may explore product-market factors, such as when vertical brand differentiation increases customer acquisition (Stahl et al. 2012), and financial-market factors, such as when vertical brand differentiation lowers the cost of capital (Larkin 2013).
Limitations
There are limitations in our data. The sample of firms that apply to the Fortune list is composed of mostly large, well-performing, publicly traded firms. Even though we control for potential selection bias, the effects we document may not extend to smaller or privately held firms. Moreover, given the nature of our employee and pay data, our results only apply to the most representative employees of these firms. Finally, we do not have a direct measure of employee differentiation in our archival data. While we use experiments with HR managers to manipulate employee differentiation to offer evidence about our effects, future research could use more fine-grained employee data to test our ideas more completely.
Conclusion
We uncover a unique role for brands in labor markets. Distinguishing between vertical and horizontal sources of brand differentiation, we find that vertical (horizontal) brand differentiation is associated with lower (higher) pay and that firms high in horizontal brand differentiation pay more for employees who offer brand-relevant complementarities. Our findings show that these brand–pay relationships have important downstream effects. Specifically, when managers leverage vertical brand differentiation to lower pay, profits are lower due to the mediating effects of lower employee productivity and retention, and that when managers pay more at firms high on horizontal brand differentiation, profits are higher via the same mediating employee behaviors. We also identify a set of firm strategies and investments in the employee–brand matching process that moderate the brand–pay relationship and downstream profit effects.
Supplemental Material
sj-pdf-1-mrj-10.1177_00222437231184429 - Supplemental material for Brands in the Labor Market: How Vertical and Horizontal Brand Differentiation Impact Pay and Profits Through Employee–Brand Matching
Supplemental material, sj-pdf-1-mrj-10.1177_00222437231184429 for Brands in the Labor Market: How Vertical and Horizontal Brand Differentiation Impact Pay and Profits Through Employee–Brand Matching by Christine Moorman, Alina Sorescu and Nader T. Tavassoli in Journal of Marketing Research
Footnotes
Acknowledgments
The authors thank BAV Consulting, a Y&R Brands company, for providing the brand data and to the Great Place to Work Institute for the employee data used in the archival research. Special thanks to Dan Cable, Rajesh Chandy, Gavan Fitzsimons, Manpreet Gill, Harald van Heerde, Mark Houston, Rik Pieters, Mengze Shi, Stephen Spiller, Michel Wedel, and seminar participants at Duke University, Florida State University, INSEAD, KU Leuven, London Business School, Melbourne Business School, Monash University, North Carolina State University, University of Houston, University of Maryland, University of Miami, UNSW Business School, University of South Carolina, University of Utah, University of Wisconsin–Madison, and VU University for their feedback on an earlier draft or presentation of this manuscript. Thanks also to Rodrigo Diaz, Holly Howe, and Licheng “Mike” Sun for assistance in running studies and to the members of chapters of the Society for Human Resource Managers for participating in their studies.
Authors Contributions
The authors are listed in alphabetical order to reflect equal contributions.
Coeditor
Vikas Mittal
Associate Editor
Alok Saboo
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
