Abstract
Long-term customer relationships develop over repeated interactions, underscoring the importance of frontline employees (FLEs) engaging in ethical behaviors. Therefore, organizations must understand how a strong ethical climate (EC) may affect attitudes and behaviors among FLEs. This study reviews frontline-related EC research and employs a meta-analytic approach to investigate the direct, indirect, and contingent effects of EC on FLE actions, attitudes, and outcomes. The authors reviewed 67 frontline-related studies comprising a sample of 21,118 respondents to assess meta-analytic associations and derive a model for structural testing. The findings from this study show that a strong EC drives customer-oriented behaviors, fosters desirable job attitudes, reduces felt stress, increases perceived performance, and decreases turnover intentions among FLEs. The strength of theses associations is often predicated on individual-level (FLE experience), study-level (response rate), and country-level (perceived corruption, individualism/collectivism) factors. This study offers theoretical and managerial contributions germane to multiple uncertainties in service literature about EC’s implications on FLEs, including EC’s ability to break through sources of tension-facing FLEs, the mediated nature of EC’s impact on perceived performance through frontline actions, and the generalizability of the economic and human benefits of EC across service contexts and frontline roles that foster greater diffusion in practice.
As the complexity of business exchange increases, so too does the importance of institutional alignment (Hartmann, Wieland, and Vargo 2018) and the relevance of developing an ethical workforce (Agnihotri et al. 2012). Relatedly, research also notes the heightened significance of establishing norms of behavior that help guide employees with the organization’s integrity in mind within settings in which employee behaviors are undertaken in a customer-facing manner (Mulki, Jaramillo, and Locander 2006). Thus, ethical climate (EC) is especially relevant to service contexts because frontline employees (FLEs) are in direct contact with customers (DeConinck, DeConinck, and Banerjee 2013). However, while research argues that a linkage exists between FLE ethical behavior and organizational performance (Schwepker and Hartline 2005), the extent to which organizations buy-in to this linkage is an open question. Ethical scandals have become a common occurrence in service contexts (von Walter, Wentzel, and Tomczak 2016), and organizations continue to engage in deceptive practices at the expense of their customers and reputation (Karpen, Bove, and Lukas 2012). Consider, for example, service firms popularized in the press for enabling a culture of miscreant actions at the top (e.g., Uber) and deceitful practices on the frontlines (e.g., Wells Fargo; Lublin 2017). These examples illustrate how unethical climates can filter down into employee actions, create pervasive reputational effects, and damage FLE dispositions and performance in an enduring way.
Within service functions, the impact of EC is amplified at the frontlines because the firm’s culture is largely reflected in the behavior of FLEs (von Walter, Wentzel, and Tomczak 2016). FLEs are in direct contact with customers and charged with navigating complex situations with little governance (Schwepker and Hartline 2005). However, while FLEs’ perceptions of the firm’s EC may orient their actions within customer interfaces (Hill and Watkins 2007), FLEs also face ethical tensions and role pressures that may be at odds with each other (Bolander et al. 2017). Thus, it is unclear whether FLEs will consistently act with customers’ interests in mind simply because of the impetus of their organization’s EC (i.e., whether the impact of an EC will vary in the service context; Coelho et al. 2010; Kadic-Maglajlic et al. 2019). For these reasons, the intersection of ethics and the frontline is an important research domain (Agnihotri and Krush 2015; Ingram, LaForge, and Schwepker 2007; von Walter, Wentzel, and Tomczak 2016), and a meta-analysis is required to integrate the abundance of empirical research to resolve such questions. Thus, this study investigates the role of EC in determining FLE turnover and perceived performance through a meta-analytic investigation, incorporating several critical mediators (e.g., customer-oriented behaviors).
Collectively, this meta-analysis expands the understanding of EC’s impact on FLEs and organizationally relevant outcomes by advancing an integrated downstream EC framework. This framework investigates a variety of direct, indirect, and contingent relationships extending from EC that offer important advancements to the literature. While established literature broadly organizes EC’s implications along the lines of employee actions, attitudes, and outcomes (e.g., DeConinck 2011; Treviño, Butterfield, and McCabe 1998), the majority of studies in the area assess its influence in piecemeal form. As a result, EC’s role within its broader nomological network is not well defined. Extant research is in further need of integration because the effects occurring throughout the network of relationships surrounding EC are moderated by several boundary conditions. As such, this framework provides an evidence-based understanding of EC that can be extrapolated across service contexts (see Figure 1).

Ethical climate at the frontline meta-analytic framework.
The findings from this study show that a strong EC fosters FLEs’ desirable job attitudes, reduces felt stress, and decreases turnover intentions. These findings underscore the need for researchers to examine EC among FLEs (Evans et al. 2012; O. C. Ferrell, Johnston, and Ferrell 2007), as well as to gain insights into the costly impact of FLE turnover on customer service quality (Ashill, Rod, and Carruthers 2008), service performance (Kraemer, Gouthier, and Heidenreich 2017), service brand image (Subramony and Holtom 2012), and service business failure (Schlesinger and Heskett 1991). The findings also show that EC drives customer-oriented behaviors and increases FLE perceived performance, but that FLE customer orientation does not mediate the EC–performance relationship. These findings help solidify the debated relationship between EC and customer orientation and clarify uncertainties in how particular mechanisms influence desired downstream outcomes (Lau et al. 2017). Research would value understanding the association among ethics, customer orientation, and performance outcomes (Bateman and Valentine 2015; Howe, Hoffman, and Hardigree 1994); the mediated nature of this relationship (or lack thereof; DeConinck 2010); the incorporation of relational bonds as a coinciding foundation for a summative ethics study (Abela and Murphy 2008); and the generalizability of such models at the frontline (Kadic-Maglajlic et al. 2019).
In addition, this study is the first meta-analysis to understand the network of relationships surrounding EC within an FLE-focused sample. As revealed from our review of ethics-related meta-analytic inquiries (see Table 1), existing meta-analyses have explored ethics-related phenomena in the population of general employees. However, given the substantial differences between internal- and external-facing roles, research needs to examine manifestations of ethics specific to FLEs and include variables of relevance to the service domain (e.g., customer orientation). Meta-analysis provides a robust conceptual foundation that may not be discernable in a narrative review (Grewal, Puccinelli, and Monroe 2018), while also empirically providing the power to test for between-study variance and boundary conditions that reflect study design choices (Roschk, Loureiro, and Breitsohl 2017). These advantages are important to service scholarship and practice, enabling scientific understanding through the integration and generalization of key findings in a targeted research domain (Grewal, Puccinelli, and Monroe 2018). Applying these advantages, the findings from this study show that individual-level (FLE experience), study-level (response rate), and country-level (perceived corruption, individualism/collectivism) variables moderate the impact of EC on FLE outcomes.
Examples of Previous Ethics-Related Meta-Analyses.
Note. JBE = Journal of Business Ethics; JAP = Journal of Applied Psychology; JBR = Journal of Business Research; IV = independent variable; MV = mediating variable; DV = dependent variable.
a Articles reporting evaluative information on meta-analytic collection and procedures are included. Narrative reviews and scoping meta-analyses are excluded.
The collective findings of this study offer the service field enhanced understanding of how EC can foster stronger customer relationships and improve organizational outcomes. As service organizations rely on FLEs to internalize broader firm strategies and cultural values in their customer-facing role but, at the same time, empower FLEs to operate independently, the organization’s EC as a guiding force becomes ever more relevant. Frontline research reflects this notion, as it aims to investigate phenomena among the tenets of the service-dominant logic and service ecosystems, with service firms striving to make interorganizational exchange easier through institutional norms (Hartmann, Wieland, and Vargo 2018). While scholars assert the importance of managing the EC of FLEs as one such norm (e.g., Schwepker and Hartline 2005), the current study provides generalizability across service contexts (in which ethical violations may have greater/lesser effect) and FLE roles (in which greater/lesser pressures to violate ethical standards may exist). In addition, the findings affirm the value of strengthening EC to enhance FLE actions and attitudes, which in turn increase resource gains (performance) and decrease resource losses (turnover). These extensions indicate that EC can offer a source of competitive advantage in frontline contexts in which consumers must trust that their inputs will be efficaciously considered (Briggs, Jaramillo, and Weeks 2012). As these implications are unique to the service domain and the FLEs who represent service organizations, our meta-analysis adds a foundational understanding to the current body of knowledge on EC at the frontline.
Literature Review and Theoretical Framework
FLEs and Ethical Tensions
Appraising the workplace environment is important, as perceptions of concepts such as organizational climate vary within an organization depending on an individual’s work group (Victor and Cullen 1988). FLEs operate at the boundary of an organization and its customers, and their relational success depends in part on their ethical approach (Hansen and Riggle 2009). However, FLEs are targets of ethical criticism and susceptible to unethical behavior given the “intense pressure to meet their performance goals” (Bolander et al. 2017, p. 519).
Throughout all elements of FLEs’ customer-facing role, understanding their environmental surroundings is vital to gain a holistic perspective on which factors may predispose them to act unethically. For example, FLE role demands often lead to increases in unethical behavior (Pettijohn, Pettijohn, and Taylor 2008). More specifically, role pressures (e.g., performance pressure, financial pressure, intraorganizational pressure, and extra-organizational pressure) and role conditions (e.g., boundary-spanning role, absence of observable supervision, rejection rates, and information asymmetry) can affect the extent of ethical tension FLEs face. These conditions summarize the contexts in which FLEs operate that potentially trigger misdirected actions, unfavorable attitudes, and diminished performance, as well as the need for service firms to strengthen their EC as a governance mechanism for their FLEs (Schwepker and Hartline 2005).
FLEs and EC
With growing recognition of ethics-related business problems, researchers have increasingly focused on the work environment (Agnihotri et al. 2012; Gustafson, Pomirleanu, and John-Mariadoss 2018). As a mechanism to alleviate ethical pressures imposed on FLEs, organizations can develop a strong EC to foster correlating acts among employees (Hunt, Wood, and Chonko 1989; Ingram, LaForge, and Schwepker 2007). EC is “a composite of organizational members’ perceptions of the ethical values and behaviors supported and practiced by organizational members” (Schwepker and Good 2004, p. 170). EC encompasses employee perceptions of the organization’s ethical rules, policies, procedures, and norms that shape acceptable workplace behaviors (Martin and Cullen 2006; Mulki, Jaramillo, and Locander 2008). Multiple theoretical frameworks reinforce EC’s role in shaping FLE actions and attitudes, including organizational socialization (Goebel and Weißenberger 2017; Ruiz-Palomino and Martínez-Cañas 2014), role stress (Mulki, Jaramillo, and Locander 2008), social contracts (Itani, Jaramillo, and Chonko 2019), and organizational identification (Briggs, Jaramillo, and Weeks 2012).
The domain of interest is particularly relevant to the boundary-spanning FLE subculture that aims to meet customer expectations and develop long-term relationships (O. C. Ferrell, Johnston, and Ferrell 2007; Itani, Jaramillo, and Chonko 2019; Mulki, Jaramillo, and Locander 2008). In other words, a firm’s EC is of particular interest to FLEs because it drives values, energizes behaviors, and influences ethics (Verbeke, Ouwerkerk, and Peelen 1996). However, most studies in the area assess EC’s influence on such outcomes in an isolated manner, thus creating a need for an integrated EC framework, particularly within the service domain (Schwepker and Hartline 2005).
EC and the Service Domain
Frontline activities are perhaps best understood in terms of “service-for-service exchange,” a focal principle to the notion of the service-dominant logic (Vargo and Lusch 2017). This logic extols the primacy of service and value creation, which increasingly occur within a broad system of actors who support frontline activities (Vargo and Lusch 2016). Such systems of actors include the service ecosystem perspective of service-dominant logic, defined as “relatively self-contained, self-adjusting system[s] of resource-integrating actors connected by shared institutional arrangements and mutual value creation through service exchange” (Vargo and Lusch 2016, pp. 10-11). From this perspective, frontline activities create and maintain crossing points with the exchange partner—with “thin” crossing points reflecting alignment on what is being exchanged and on institutional arrangements and “thick” crossing points reflecting unclear or misaligned norms and representations across actors (Hartmann, Wieland, and Vargo 2018). While these crossing points and service-for-service exchange behaviors are interrelated, they are not fully exemplified independent of broader structural influences (Vargo and Lusch 2016).
Hartmann, Wieland, and Vargo (2018) thus contend that the service ecosystem perspective offers a strong theoretical framework for examining frontline service-related phenomena—one that expands the view from dyadic exchanges to broader practices influenced by institutional systems. In relation to this meta-analysis, there is value in taking a step back from solely assessing FLE and customer exchanges and adopting a view that also includes systems-level phenomena that may shape customer value creation practices (e.g., EC). The service ecosystem positioning does not diminish the importance of the frontline in service exchange but rather suggests that considering elements from other levels of aggregation (i.e., cultures in which dyads are embedded) can add to the collective understanding (Chandler and Vargo 2011).
Service organizations aim to increase institutional alignment to decrease the cost of exchange (i.e., thin crossing points) and avoid misalignments that may prevent exchange (i.e., thick crossing points). We posit that one means of creating institutional alignment (i.e., resolving tensions, supporting commonalities; Zietsma and McKnight 2009) is through the portrayal of a strong EC within the service ecosystem. While frontline activities are microlevel processes that can facilitate service exchange, they reflect organizational cultures and institutional work processes that shape value creation practices across time (Hartmann, Wieland, and Vargo 2018; Vargo and Lusch 2016). Because value creation takes place in service ecosystems, we expect that as service organizations strengthen their EC and FLEs act in a manner that reflects their organization’s values, institutional alignment will increase and crossing points will become thinner, thus generating desirable downstream outcomes for organizations and the individuals involved. Given the implications of EC and FLEs within service ecosystems, a targeted meta-analysis provides an important foundational understanding for the field.
EC’s Impact on Frontline Actions, Attitudes, and Outcomes
EC and Frontline Actions
EC increases FLE customer-oriented behaviors
Organizations are mobilizing customer-oriented behaviors to create stronger customer relationships and more sustainable competitive advantages in the marketplace (Bateman and Valentine 2015; Hill and Watkins 2007). Coelho et al. (2010, p. 1344) define customer orientation as follows: “A customer-oriented [FLE] directs his/her efforts towards serving the interests of the customer, discovering the customer’s needs before developing or presenting the appropriate solution, while avoiding the utilization of manipulative, deceptive, and pressure tactics.” The process of understanding organizational practices that encourage FLEs to adopt customer-oriented behaviors, however, is a challenge for many organizations (Kennedy, Lassk, and Goolsby 2002; Thakor and Joshi 2005). One suggestion indicates that creating appropriate organizational values is essential to building a customer-oriented workforce (Coelho et al. 2010; Hartline, Maxham, and McKee 2000). Related research shows the effect of FLEs’ perceptions of their own ethical treatment on customer-oriented outcomes (Hansen and Riggle 2009) and the ability of a moral atmosphere to encourage stronger customer relationships (Murphy, Laczniak, and Wood 2007). Thus, despite a critical disconnect in the literature between EC and customer orientation (Gustafson, Pomirleanu, and John-Mariadoss 2018), research builds the case for developing an ethical culture to motivate FLEs to use more relational approaches with their customers (Bateman and Valentine 2015).
The EC of an organization should increase an FLE’s customer-oriented behaviors (Swanson, Kelly, and Dorsch 1998). Research shows that FLEs who work in organizations with a strong EC are more altruistic toward others (Itani, Jaramillo, and Chonko 2019) and that FLEs’ customer-oriented behaviors directly reflect the values of their organizations (Howe, Hoffman, and Hardigree 1994). Certain customer-oriented behaviors, such as showing concern for clients and doing what is right for them, may be associated with EC (Wimbush, Shepard, and Markham 1997). Thus, FLEs may be more likely to undertake customer-oriented acts when they perceive their work environments as ethical (Coelho et al. 2010). Given such evidence and the notion that an “[EC] presents an added barrier to unethical employee actions” (Babin, Boles, and Robin 2000, p. 345), we hypothesize the following:
EC and Frontline Attitudes
EC increases desirable FLE job attitudes
FLEs’ attitudes are influenced by how positive they feel about their business environment (Pettijohn, Pettijohn, and Taylor 2008); thus, ethical organizations are well positioned to effectively manage desirable FLE job attitudes because of their attention to maintaining fairness (Valentine, Fleischman, and Godkin 2015). Desirable job attitudes “reflect [employees’] sense of connection or commitment to their work or to their organization” (Davis and Rothstein 2006, p. 409) and consist of FLE job satisfaction and organizational commitment and identification (common attitudinal assessments to an EC). FLEs who work for organizations that maintain high ECs will have correspondingly high scores of desirable job attitudes including job satisfaction (e.g., Jaramillo, Mulki, and Solomon 2006; Schwepker 2001; Weeks and Nantel 1992) and organizational commitment (e.g., Hunt, Wood, and Chonko 1989; Kelley and Dorsch 1991).
The relationship between ECs and desirable FLE job attitudes is supported by myriad explanatory mechanisms suggesting that when ethical standards are in place, employees develop favorable attitudes toward their job and organization (Treviño, Butterfield, and McCabe 2001; Valentine and Barnett 2003). Underlying these relationships is the belief that FLEs in firms with strong ethical guidance feel more comfort and assurance in their jobs (Amyx et al. 2008; Mulki, Jaramillo, and Locander 2008, 2009). The relationship is also justified at a broader level, as FLEs who perceive their organizations as engaging in ethical practices have positive dispositions toward their jobs and take pride in their work (Pettijohn, Pettijohn, and Taylor 2008). Arguments affirming this relationship submit that employees prefer to work in organizations with positive ECs (Valentine and Barnett 2003), and ethical incongruities can lead to weaker attitudinal commitment (Levy and Dubinsky 1983). Thus, we hypothesize the following:
EC decreases FLE felt stress
Felt stress is a construct embodying tensions that FLEs often experienced in their boundary-spanning function in the organization. Organizations strive to minimize controllable elements of FLE stress to reduce its downstream consequences (Behrman and Perreault 1984; Boles, Wood, and Johnson 2003). One such mechanism for reducing felt stress is to develop, facilitate, and enforce a strong EC (Mulki, Jaramillo, and Locander 2008; Schwepker and Hartline 2005).
As most stress models rely on a stressor–strain framework, in which stressors affect work conditions and the strain on employee emotions that follow (Jex et al. 2001), EC can provide a positive work condition that results in lower FLE stress (Mulki, Lassk, and Jaramillo 2008). The more FLEs perceive the organizations as ethical, the less likely they are to assess work conditions as stressful (Babin, Boles, and Robin 2000). FLEs are better equipped to deal with stressors when they perceive their ECs to be strong because they know their organizations will reliably do what is right. Therefore, in line with extant research that suggests FLEs’ perceptions of EC directly affect their work-related stress (Jaramillo, Mulki, and Solomon 2006; Mulki, Jaramillo, and Locander 2008), we hypothesize the following:
EC and Frontline Outcomes
EC increases FLE perceived performance
One of the intended benefits of developing a strong EC is that FLEs become better able to achieve their job performance goals (i.e., key metrics and expectations associated with the customer-facing role; Weeks and Nantel 1992). However, researchers have raised the question, is good ethics good for business (e.g., Babin, Boles, and Robin 2000; Itani, Jaramillo, and Chonko 2019; Weeks et al. 2004)? To this end, research underscores the linkage between ethics-related variables and FLE performance dimensions (e.g., Agnihotri and Krush 2015; Babin, Boles, and Robin 2000; Kelley and Dorsch 1991), generally confirming the argument that FLEs’ ability to portray their organizations favorably improves FLE performance (Briggs, Jaramillo, and Weeks 2012).
A strong EC helps the organization develop mutually beneficial relationships with both customers and FLEs (Evans et al. 2012). According to Karatepe (2013, p. 77), “employees with favorable perceptions of the firm’s [EC] are embedded in their jobs. Such employees in turn display elevated levels of job performance and extra-role customer service behaviors.” On the customer side of this association, as the supplier’s ethical reputation is established, customer satisfaction, loyalty, retention, and repurchase intention are enhanced (Mulki and Jaramillo 2011). On the FLE side, FLEs who work in organizations that have well-established ethical policies are more apt to achieve job performance (Weeks and Nantel 1992). This impact on job performance is inclusive of various sales performance dimensions, such as reaching objectives and behavioral targets and mitigating unnecessary expenses to reach these goals (Jaramillo, Mulki, and Solomon 2006; Schwepker and Ingram 1996), and service performance dimensions, such as service quality and the efficiency and effectiveness of service (Babin, Boles, and Robin 2000; Luria and Yagil 2008; Weeks et al. 2004). As such, we hypothesize the following:
Research also suggests that EC is indirectly linked to FLE performance (Gustafson, Pomirleanu, and John-Mariadoss 2018; Mulki, Jaramillo, and Locander 2009), primarily through its influence on FLE actions and attitudes, which can facilitate desired downstream effects on performance (Karatepe 2013; Schwepker and Good 2011; Weeks et al. 2004). To the extent that EC filters into customer loyalty and positive dispositions toward the supplier (Hansen and Riggle 2009; Mulki and Jaramillo 2011), customer-oriented behaviors act as a conduit for such relational performances to take hold. An EC influences FLE prioritization of customer interests, which should subsequently improve performance (Babin, Boles, and Robin 2000; Luria and Yagil 2008). With regard to both FLEs’ actions and attitudes, several procedural mechanisms can explain the mediated relationships: (1) EC encourages ethical behaviors (Schneider and Bowen 1985) and has a positive impact on attitudes toward fairness (Babin, Boles, and Robin 2000) and (2) ethical behaviors (Román and Munuera 2005; Schwepker and Ingram 1996; Tanner, Tanner, and Wakefield 2015) and desirable attitudes (Simons and Roberson 2003) positively affect performance. In addition, Mulki, Jaramillo, and Locander (2008) show that EC reduces emotional exhaustion because it clarifies ethical expectations. FLEs who experience a lower level of strain are more productive and perform at a higher level (e.g., Low et al. 2001). Taken together, a positive EC is likely to promote customer-oriented actions, drive desirable job attitudes in FLEs, and reduce felt stress, ultimately increasing FLE performance (DeConinck 2010; Mulki and Jaramillo 2011; Mulki, Jaramillo, and Locander 2008). Accordingly, we hypothesize a mediated relationship through FLE actions and attitudes:
EC decreases FLE turnover intention
The literature underscores the need to investigate the extent to which organizations operating at high ethical levels produce both better business results and better stakeholder outcomes (e.g., DeConinck 2010; Grisaffe and Jaramillo 2007). FLE turnover is a costly problem because employees who work on the frontline tend to exit at a high rate and are expensive to replace (Boles et al. 2012). While research has investigated myriad FLE factors that can decrease turnover intentions (i.e., an FLEs plan to exit the organization), studies have also encouraged a focus on conditions under managerial control (Skiba, Saini, and Friend 2016). With the ability to directly and indirectly influence FLE turnover intentions, EC represents one such factor.
When FLEs feel pressure from the EC to engage in questionable tactics that are personally unsuitable to them, increased turnover may result (Fournier et al. 2010). Thus, FLEs tend to leave their employers at higher rates when working in environments they consider ethically challenging (DeConinck 2010; Levy and Dubinsky 1983; Mulki, Jaramillo, and Locander 2008). Conversely, individuals tend to remain with organizations they perceive as ethical (Schwepker 2001). We therefore hypothesize the following:
Beyond this direct effect, however, EC is also likely to reduce turnover through its positive association with FLE actions and attitudes. An EC creates a work environment in which FLEs feel satisfied with their jobs and committed to their organizations and thus are less likely to quit (Mulki, Jaramillo, and Locander 2006). Multiple theoretical explanations of the relationship between EC and turnover support the partially mediated relationship through FLE attitudes. First, attitudinal theory suggests that individual evaluations of an object lead to attitudes that subsequently explain behavioral intentions (Ajzen 2001), thus supporting a trickle-down effect of a firm’s EC on higher job fulfillment and lower turnover intentions (Mulki, Jaramillo, and Locander 2006, 2008; Schwepker 2001). Second, theory also purports that lower levels of stress lead to lower levels of turnover (Rizzo, House, and Lirtzman 1970); thus, EC’s ability to provide cues about appropriate behaviors can reduce felt stress and, thereby, turnover intentions (Schwepker and Hartline 2005). With regard to FLE actions, EC’s ability to promote customer-oriented behaviors helps decrease turnover (McFarland 2003). Extant research collectively reinforces these mediated relationships between EC and turnover through FLE actions and attitudes (e.g., Jaramillo, Mulki, and Solomon 2006; Mulki, Jaramillo, and Locander 2006; Schwepker 2001), leading us to hypothesize the following:
Moderating Effects on EC
As depicted thus far, the EC of an organization is likely to influence a variety of FLE actions, attitudes, and outcomes. However, the direct effect between EC and the FLE is not universally supported (see Kadic-Maglajlic et al. 2019). Rather, (1) FLEs bring a variety of individual-level experiences to each situation they encounter, which alters their evaluative lenses; (2) various methodological approaches result in varied response rates, which influences the nature of the final sample; and (3) country-level conditions influence broader ethical perceptions of respondents. This study clarifies such ambiguities by assessing the conditions under which EC is more or less likely to produce an outcome, specifically by testing the moderating effects of individual-level (FLE experience), study-level (response rate), and country-level (perceived corruption, individualism/collectivism) moderators within the theoretical model. 1
FLE experience
An EC can be both a barrier to unethical employee actions (Babin, Boles, and Robin 2000) and a conduit for desirable attitudes toward the job and organization (Treviño, Butterfield, and McCabe 2001; Valentine and Barnett 2003). Novice FLEs have less exposure to their organizations’ EC, and thus EC is less likely to influence their actions, attitudes, and outcomes. By contrast, longer tenures influence FLEs’ ethical decision-making (L. Ferrell and Ferrell 2009), and experienced FLEs possess more skills with which to identify and react to customer goals (Franke and Park 2006; Jaramillo et al. 2009a) and develop desirable job attitudes (Cicala et al. 2014). Thus, FLE experience and EC are likely to have complementary interaction effects. Specifically, while new FLEs may be able to assess their organizations’ EC, this climate may not yet have had enough time to affect their actions, attitudes, and outcomes. Conversely, long-tenured FLEs have worked within their climates for longer periods, providing greater opportunity for ECs to influence these outcomes. This argument finds support in research showing that experienced employees exert greater reciprocity effects on desirable FLE outcomes stemming from elongated organization–FLE alignment and socialization and that, given their increased exposure to ethical dilemmas, they are more willing to accept ethical standards and behave accordingly (Serwinek 1992). Thus, we hypothesize that the effect of EC on FLE actions, attitudes, and outcomes will be stronger among FLEs with greater experience:
Study response rate
Meta-analyses in the EC domain have not assessed whether the level of response rates achieved in the sampled data collections affects EC’s downstream associations (see Table 1). However, surveys may be polarized from the U-shaped tendency of those being most vocal when showing a strong positive or negative valence, in contrast with a neutral or mild opinion (Anderson 1998). As such, when survey response rates are low, the sample may contain more polarized individuals who have a generally positive or negative view of their work situations. This generalized valence will then moderate the strength of the associations between EC and its outcomes (i.e., more high–high and low–low pairings in the data between EC and positively valenced constructs; more high–low and low–high pairings between EC and negatively valenced constructs). Accordingly, samples with low response rates may be overestimating the relationships between EC and its predicted outcomes, as they are “missing the middle” of respondents with less-valenced appraisals of their work situation. However, as response rates increase, more of these moderate respondents are included in the collection, resulting in a more diversified set of respondents, including those who may be less likely to be globally positive or globally negative about their organizational life. Such samples are thus more likely to weaken associations between variables such as EC and its outcomes. Therefore, as response rates increase, the ratio of polarized to nonpolarized FLEs decreases and the relationships between EC and its outcomes become attenuated. Thus:
Country-level corruption perception
The issue of global business ethics and corruption is often regarded as “the ultimate dilemma” for global corporations (Wilhelm 2002, p. 178). Corruption increases the cost of doing business and reduces foreign direct investment (Cuervo-Cazurra 2006). Country-level corruption is typically assessed with the Corruption Perception Index (CPI), administered by Transparency International. Corruption refers to “the abuse of entrusted power for private gain” (Transparency International 2018). Research shows that the CPI affects an array of macroeconomic variables, including gross domestic product, risk perceptions, foreign investment, and trade (e.g., Davis and Ruhe 2003; Wilhelm 2002). Research also indicates that the CPI affects individual perceptions about ethics. In countries with a high level of corruption, individuals are more likely to believe that they need to compromise ethical principles to succeed in business (Jaffe and Tsimerman 2005) and be more lenient in evaluating unethical actions from service providers (Ahmed, Chung, and Eichenseher 2003). In other words, when perceived country-level corruption is high, this macro-level factor likely mutes the impact of an organization’s EC, due to the extent to which these ethical norms are deeply entrenched in a populous. As such, we propose that EC has less of an impact on FLE actions, attitudes, and outcomes when perceived country-level corruption is high. Conversely, we propose that EC has more of an impact when perceived corruption is low, representing a synergistic macro- and microlevel impact of mutually ethical environments. Thus,
Country-level individualism/collectivism
Among Hofstede’s (2001) core national culture dimensions, research has emphasized individualism/collectivism as prominently affecting ethical influences because of its focus on individual priorities versus group interests (Li and Murphy 2012; Steenkamp and Geyskens 2006). Conceptually, individualist cultures tend to have loose group ties—individuals look after their own interests, derive identity from the self (vs. organizations), and value individual expression (Blodgett et al. 2001; Hofstede 2001). By contrast, collectivist cultures (i.e., low individualism) tend to have a strong sense of belonging and reciprocity—individuals define themselves in terms of the group, view themselves as a member of an organization, and prioritize work-unit solidarity (Blodgett et al. 2001; Hofstede 2001). Frontline research has emphasized the relevance of this cultural dimension (e.g., Blut, Wang, and Schoefer 2016). Scholars also note that an FLE in a low individualism context is more influenced by group norms and less likely to behave in a manner that negatively affects the group; “strong cohesiveness in groups is the norm, and salespeople can readily turn to each other for help” (Li and Murphy 2012, p. 231). As such, FLEs in a high individualism context are less considerate of such norms and are more likely to bend rules to achieve personal gains (Blodgett et al. 2001; Vitell, Nwachukwu, and Barnes 1993). Extant theorizing thus suggests that when country-level individualism is higher, the influence of EC on FLE actions, attitudes, and outcomes is lower. Specifically, we hypothesize the following:
Method
We followed meta-analytic process recommendations as specified in recent meta-analytical and methodological research (e.g., Grewal, Puccinelli, and Monroe 2018; Johnson and Jaramillo 2017). We first clearly delineated the study’s research purpose—to investigate EC’s influence on FLEs. In this pursuit, we began by using representations from extant literature to conceptualize the focal construct of EC. We then reviewed the EC literature from its inception in the 1980s and identified the downstream variables of customer-orientated behaviors, desirable job attitudes, felt stress, perceived performance, and turnover intentions.
Database Development
We obtained studies for meta-analytic testing using multiple approaches. We sought published and unpublished research containing effect sizes involving relationships germane to our inquiry of EC and FLEs. For this search, we carried out electronic searches in Google Scholar, Business Source Complete, and Emerald. The electronic search was completed in late 2018 and included several variants of search strings to identify candidate articles in the service domain (e.g., ethical climate service, ethical climate frontline, ethical climate sales, ethical management service, and unethical corporate climate service). We used standard querying conventions and variants of included terms (e.g., frontline and front line) to maximize results. We augmented our search by using both backward- and forward-looking reference searches (Johnson and Jaramillo 2017). We sought articles providing a narrative review of the domain to generate article candidates in previous research (Lipsey and Wilson 2001). We also applied forward-looking techniques such as identifying scale development work and examining articles subsequently citing such work.
Consistent with meta-analytic best practices (see Johnson and Jaramillo 2017), we also conducted manual journal searches of Journal of the Academy of Marketing Science, Journal of Business Ethics, Journal of Marketing, Journal of Marketing Research, Journal of Personal Selling & Sales Management, and Journal of Service Research to ensure applicable articles were collected. To identify unpublished studies, we searched ProQuest’s dissertations and posted a call for relevant studies on the Electronic Marketing List Information listserve.
Study eligibility was limited to articles reporting a Pearson’s correlation (r) or other statistics that could be converted to r (e.g., p values, t statistics, χ2). Studies using an organizational-level construal of ethics were included, while studies exploring individual-level ethical phenomena (i.e., ethical behavior) were excluded. In addition, consistent with this study’s aim to advance meta-analytic insights specific to FLEs and FLE-relevant variables, we omitted studies using general employees.
Data Coding
We developed a detailed codebook for use on the studies including empirical relationships between EC and downstream FLE variables (Grewal, Puccinelli, and Monroe 2018). Table 2 provides the names and definitions of each construct, along with coding elements used in the analysis. The table also contains representative articles, common aliases for each measure, and exemplar items from established scales.
Names, Definitions, Representative Articles, Common Aliases, and Example Items for Variables Included in the Meta-Analysis.
Note. EC = ethical climate.
a r = Reverse-coded item. bCPI from Transparency International (https://www.transparency.org/cpi2018). cHofstede Individualism Index (https://www.hofstede-insights.com/product/compare-countries/insights.com/product/compare-countries).
We manually coded the data. This process included specifying the relationships included and their bivariate coefficients, as well as delineating study-specific factors (e.g., sample size, response rate, construct reliabilities, experience, and country of origin) for use in subsequent analyses. To assess intercoder reliability, two of the authors coded the same subset of representative articles on the various main effects and moderating elements examined in the meta-analysis for a subsample of 25 articles. The results revealed high levels of convergence (93.6%) between the authors on the coded data. All differences in coding were resolved through discussion. We also verified that all studies included in the database were independent by comparing studies with similar authors and sample sizes. This process rendered 67 studies from 66 articles and 21,118 respondents. The data came from 21 different countries, sample sizes ranged from 24 to 1,455 respondents, responses came from relatively experienced FLEs (average of 10.2 years), and response rates ranged from low (11%) to complete participation (100%; see also Online Appendix).
Meta-Analytic Procedures
We transformed raw correlations into Fisher-Z coefficients and weighted them by their inverse variance (Lipsey and Wilson 2001). We meta-analyzed the Fisher-Z transformed correlations using a random-effects model in Metawin Version 2.0. The random-effects model accounts for possible heterogeneity in the population effect size, thus rendering more conservative and generalizable results (Hedges and Vevea 1998). Raw correlations (r) were disattenuated (rc ) to account for measurement error (Goad and Jaramillo 2014). Table 3 reports the mean effect sizes and 95% confidence intervals. We assessed homogeneity of the bivariate relationships with Q statistics (see Rubera and Kirca 2012). In addition, we calculated I 2 statistics to estimate the proportion due to heterogeneity (Higgins and Thompson 2002). Both statistics suggest that a significant level of heterogeneity can be explained by moderating factors (Table 3). Finally, to address publication bias, we estimated Orwin’s (1983) fail-safe N statistics. Fail-safe N represents the number of studies reporting a nonsignificant correlation (r = 0) that is necessary to reduce the meta-analytic effect size to r = .1. We selected a cutoff value of r = .1 following the conventional standard of a small correlation of practical significance (Cohen and Cohen 1983).
Derived Meta-Analytic Correlations.
a Number of independent effect sizes. bTotal sample. cAverage correlation, weighted for sample size. dAverage correlations corrected for measurement error. eLower bound, 95% confidence interval (CI) of rc . fUpper bound (UB), 95% CI of rc . g Q statistic measure of heterogeneity (fixed-effects model). h I 2 statistic (percentage of true heterogeneity to total variance). iFail-safe N = Number of effect sizes with r = 0 to render a “small” cumulative effect size (r = |.10|). jControl success rate (CSR) = 0.5 − rc /2. kSuccess rate (SR) = 0.5 + rc /2.
*p < .05.
We used meta-analytic structural equation models (SEMs) and meta-analytic regressions to assess the proposed conceptualization (e.g., Zablah et al. 2012). For SEM analyses, we also combined the meta-analytic correlation coefficients from this study (Table 3) with effect sizes obtained from previous meta-analyses (Table 4; Butts, Casper, and Yang 2013). For example, we derived the correlation involving FLE attitude and customer orientation from Zablah et al. (2012). The off-diagonal entries in the lower left of Table 4 contain the disattenuated meta-analytic correlations used to test the SEM.
Meta-Analytic Correlations Used to Test SEM.
Note. Off-diagonal entries in the lower left contain disattenuated meta-analytic correlations (r). Off-diagonal entries in the upper right show the sample sizes (Ns in parentheses) and independent effect sizes (k). Correlations with footnotes were derived from published meta-analysis. EC = ethical climate; SEM = structural equation model.
a Statistics derived from Zablah et al.’s (2012) meta-analysis. bStatistics derived from Alarcon’s (2011) meta-analysis. cStatistics derived from Goad and Jaramillo’s (2014) meta-analysis. dStatistics derived from Verbeke, Dietz, and Verwaal’s (2011) meta-analysis. eStatistics derived from Franke and Park’s (2006) meta-analysis. fStatistics derived from Griffeth, Hom, and Gaertner’s (2000) meta-analysis.
We used the harmonic mean of the sample sizes of the bivariate correlations as the sample size for the SEM (n = 4,497). Each relationship included in the study contained data from at least five samples (k ≥ 5), exceeding the k ≥ 3 criterion used in most causal models in marketing (Grewal, Puccinelli, and Monroe 2018, p. 23). Prior meta-analyses have shown that FLE experience is related to sales attitudes and behaviors (e.g., Goad and Jaramillo 2014). In view of such inclusion, we treat experience as a control variable in the SEM.
Meta-Analytic Results
Bivariate Analyses
Bivariate analyses reported in Table 3 show that EC is positively related to FLE customer-oriented behaviors (rc = .39, p < .05), desirable job attitudes (rc = .46, p < .05), and perceived performance (rc = .27, p < .05). EC is negatively related to felt stress (rc = −.21, p < .05) and turnover intention (rc = −.34, p < .05). The hypothesis results related to these associations appear in the next section. Meta-analytic correlations involving experience and EC are not significant.
SEM
A meta-analytic SEM served to test the direct effect and mediation hypotheses (Viswesvaran and Ones 1995). The proposed meta-analytic model suggested a close fit to the data (χ2 = .45, df = 1, p < .05, [Adjusted Goodness of Fit] AGFI = .999, comparative fit index = 1.0, root mean square error of approximation = .000). Table 5 provides the standardized path coefficients, t values, and direct and indirect effects. Results shown in Table 5 were used to test the direct effect hypotheses. As expected, EC increased customer-oriented behaviors (Hypothesis 1: β = .39, p < .01), desirable job attitudes (Hypothesis 2: β = .33, p < .01), and perceived performance (Hypothesis 4a: β = .19, p < .01). EC also reduced felt stress (Hypothesis 3: β = −.11, p < .01) and turnover intentions (Hypothesis 5a: β = −.09, p < .01). The results of mediation tests (Iacobucci, Saldanha, and Deng 2007) show that EC affects perceived performance through desirable job attitudes (Hypothesis 4c: Sobel test = 7.06, p < .01) and felt stress (Hypothesis 4d: Sobel test = 4.09, p < .01). EC influences turnover through customer-oriented behaviors (Hypothesis 5b: Sobel test = 4.95, p < .01), desirable job attitudes (Hypothesis 5c: Sobel test = 4.93, p < .01), and felt stress (Hypothesis 5d: Sobel test = 6.08, p < .05). The mediating impact of customer-oriented behaviors on the relationship between EC and performance was not significant (Hypothesis 4b: Sobel test = .95, p = .34). Figure 2 provides a summary of the mediation results, including paths, estimated coefficients, and significance.
SEM Results.
Note. EC = ethical climate; SEM = structural equation model.

Mediation analysis results. Supported hypotheses in bold. Hypotheses 1–5 tested with path analysis. Hypotheses 6–9 tested with meta-analytic regression.
Meta-Analytic Regressions
Meta-analytic weighted least square regression models served to evaluate the moderating role of FLE experience (Lipsey and Wilson 2001). In meta-analytic regressions, the effect size is treated as the dependent variable and the moderator as the independent variable (effect size = β0 + β1 × Moderator + ∊). We used the β1 parameter in this regression and its significance (t value) to assess moderation in a sequential manner. The results show that the positive influence of EC on customer-oriented behaviors (Hypothesis 6a: β1 = .02, p < .01) and desirable job attitudes (Hypothesis 6b: β1 = .05, p < .01) is stronger when FLE experience is high. Experience also increases the positive impact of EC on perceived performance (Hypothesis 6c: β1 = .01, p < .01) as well as the mitigating effects on turnover intention (Hypothesis 6d: β1 = −.09, p < .01).
The results also show that the positive influence of EC on desirable job attitudes (Hypothesis 7b: β1 = −.0011, p < .01) and perceived performance (Hypothesis 7c: β1 = −.0012, p < .05) is weaker when the study response rate is low. However, study response rates did not significantly affect the impact of EC on customer-oriented behaviors (Hypothesis 7a: β1 = .0007, p > .10) or turnover intentions (Hypothesis 7d: β1 = .0003, p > .10). The positive influence of EC on desirable job attitudes (Hypothesis 8b: β1 = .0033, p < .01) and perceived performance (Hypothesis 8c: β1 = .0041, p < .01), along with the negative impact of EC on turnover intention (Hypothesis 8d: β1 = −.0043, p < .01), is stronger when perceived country-level corruption is low. However, the positive effect of EC on customer-oriented behaviors (Hypothesis 8a: β1 = −.0025, p < .05) is weaker when perceived country-level corruption is low. Of note, we coded country-level corruption scores for each country with the commonly used CPI (Transparency International 2018). CPI has a reversed-scoring direction; thus, the results indicate that the effects of EC are more important in countries with lower levels of perceived corruption. Our sample included countries with both high levels of perceived corruption, such as Nigeria (CPI = 27), Mexico (CPI = 29), Pakistan (CPI = 32), and Ecuador (CPI = 32), and low levels of perceived corruption, such as Canada (CPI = 82), Finland (CPI = 85), and United States (CPI = 75). As expected, the average effect sizes involving EC on desirable job attitudes, perceived performance, and turnover intentions are weaker in countries where perceived corruption is high (CPI < 32, k = 7, r EC–Attitude = .36, r EC–Perf. = .07, r EC–Turnover = −.05) than in countries where perceived corruption is low (CPI > 75, k = 36, r EC–Attitude = .41, r EC–Perf. = .29, r EC–Turnover = −.28).
The findings also reveal that country-level individualism (IND) increases the effects of EC on desirable job attitudes (Hypothesis 9b: β1 = .0005, p < .05), perceived performance (Hypothesis 9c: β1 = .0031, p < .01), and turnover intentions (Hypothesis 9d: β1 = −.0009, p < .10). However, country-level individualism did not affect the impact of EC on customer-oriented behaviors (Hypothesis 9a: β1 = −.0007, p > .10). Our sample included data from highly individualist countries, such as Canada (IND = 80) and the United States (IND = 91), and highly collectivist countries, such as Ecuador (IND = 8), Pakistan (IND = 14), South Korea (IND = 18), and China (IND = 20). The average effect sizes involving EC on desirable job attitudes, perceived performance, and turnover intentions are stronger in individualist countries (IND > 80, k = 31, r EC–Attitude = .41, r EC–Perf. = .29, r EC–Turnover = −.31) than collectivist countries (IND < 20, k = 8, r EC–Attitude = .36, r EC–Perf. = .07, r EC–Turnover = −.25). This impact of country-level individualism was opposite the direction hypothesized.
Discussion
This study uses meta-analysis to understand EC’s influence in an FLE context. The findings reveal that EC fosters FLEs’ desirable job attitudes, reduces felt stress, and decreases turnover intentions. In addition, EC drives customer-oriented behaviors and increases FLE perceived performance, but customer-oriented behaviors do not mediate the EC–performance relationship. Finally, the meta-analysis shows that FLE experience, study response rate, perceived country-level corruption, and country individualism moderate several effects occurring throughout the network of relationships surrounding EC. These results provide a foundation for deriving valuable theoretical and managerial implications—each reflective of the ability of meta-analysis to enrich understanding of EC’s influence on FLEs and organizationally relevant outcomes.
Theoretical Implications
This study offers multiple theoretical contributions to the multidisciplinary service research context. As we find from the review of ethics-related meta-analytic inquiries, extant studies have not accounted for the unique role of FLEs when examining EC. Thus, as the first contribution and in fulfilling the objective of investigating EC among FLEs, this study contributes to the coinciding need for comprehensive studies on ethics and customer-centric approaches to develop relational bonds (Abela and Murphy 2008). By extending existing EC summative work into the service realm, this research depicts the impact of EC on actions unique to the FLE role.
While the FLE role itself varies across contexts (e.g., business-to-consumer vs. business-to-business, service support vs. service delivery), service scholars emphasize the importance of improving this generalizable understanding and ways to create a strong EC to foster reflective FLE behaviors (DeConinck, DeConinck, and Banerjee 2013; Schwepker and Hartline 2005). The findings from this study show that an organization’s EC has a positive impact on FLE customer-oriented behaviors, a construct disconnected from EC in the literature (Gustafson, Pomirleanu, and John-Mariadoss 2018). Furthermore, the meta-analytic results indicate that customer-oriented behaviors mediate EC’s impact on FLE attitudes and turnover intentions but not perceived performance. These combined results help identify organizational practices that encourage FLEs to adopt customer-oriented behaviors (Kennedy, Lassk, and Goolsby 2002; Thakor and Joshi 2005); the association among ethics, customer orientation, and performance outcomes (Bateman and Valentine 2015; Howe, Hoffman, and Hardigree 1994; Lau et al. 2017); and the mediated nature of these relationships (DeConinck 2010). Specifically, the findings suggest that as organizations project frontline performance implications from improvements made to their EC, the underlying mechanism by which the downstream performance effects take hold (Lau et al. 2017) is not through FLE customer-oriented behaviors; instead, the theoretical premise of service-dominant logic and service ecosystems suggests that performance implications derive from institutional alignment (Hartmann, Wieland, and Vargo 2018). A broader set of frontline actions—beyond FLE customer orientation—may reflect the organization’s institutional work processes and organizational culture, which in turn shape value creation practices and FLE efforts that produce thin crossing points with service beneficiaries.
As a second contribution, this study adopts a double-bottom-line assessment of financial performance and employee-focused outcomes. From a performance perspective, developing a strong EC within a service context can drive FLEs to act ethically with customers, which in turn strengthens relationships and firm revenues (Itani, Jaramillo, and Chonko 2019; Palmatier et al. 2006). The results also reveal employee-focused implications of an organization’s EC, showing the direct and indirect effects on FLE attitudes. In particular, these findings shed light on the mechanisms that reduce FLE turnover, a construct of high impact with regard to service quality (Ashill, Rod, and Carruthers 2008), service performance (Kraemer, Gouthier, and Heidenreich 2017), and service business failure rates (Schlesinger and Heskett 1991). Collectively, the findings show the impact of a strong EC and support the doctrine that good ethics is good business (Babin, Boles, and Robin 2000; Itani, Jaramillo, and Chonko 2019; Weeks et al. 2004), from both an economic and a human perspective. These findings are relevant in a frontline context given (1) the numerous ethical tensions and pressures facing FLEs that may confound ethical intent fostered by a service organization’s climate and (2) the questions that remain at the intersection of the ethics and frontline research domains (e.g., whether EC implications are realized across service contexts and FLE roles; Agnihotri and Krush 2015; Ingram, LaForge, and Schwepker 2007). Our meta-analytic results establish the multifaceted value of strengthening EC that can be generalized to the myriad FLE jobs in service organizations that are inherently captured in the studies sampled for this analysis.
Third, as uncertainties exist in the literature about the impact of cross-study variance, and as only a few studies have been able to provide support for the moderating influences affecting meta-analytic associations, the moderator findings from this study provide methodological and theoretical relevance. FLE experience accentuates the associations between EC and customer-oriented behavior, desirable job attitudes, perceived performance, and turnover intentions. Theoretically, the moderating effect of FLE experience indicates that the signals of an EC take time to set in and repeated exposure can facilitate its contagion effect. Next, regarding the moderating impacts of corruption and individualism, the findings provide partial support that the location of investigation matters. As frontline scholars aim to undertake more macro-level research (Cron 2017), and theoretical foundations suggest that such a perspective can further reveal institutional work processes and value creation practices (Chandler and Vargo 2011; Hartmann, Wieland, and Vargo 2018), this meta-analysis provides empirical support to such conceptual arguments. Both low levels of corruption and high levels of individualism strengthen the effect of EC on desirable job attitudes, perceived performance, and turnover intentions. A counterintuitive finding that individualism accentuates the influence of EC on these outcomes may be a function of the level of impact at which individualism/collectivism occurs. For example, if higher level factors (e.g., societal) predicate behavior to a greater extent than lower level factors (e.g., organizational), organizational-level factors such as EC are likely less influential on downstream outcomes when individualism is low. By contrast, societies high in individualism have less societal-level influence, thus making the organization’s EC more salient and influential on outcomes. This finding warrants further theoretical and empirical attention.
Finally, the finding that high study response rates attenuate EC’s effects on desirable job attitudes and perceived performance is a notable methodological contribution that highlights the need to maximize response rates, especially in frontline contexts in which response rates tend to be lower than in the broader marketing domain (Carter, Dixon, and Moncrief 2008). However, frontline research has yet to explore the effect of response rate on examined associations in a systematic manner using meta-analytic procedures. This is a shortcoming that needed to be addressed given that low response rates may result in a collection of FLEs with generally positive or generally negative appraisals of their job and work conditions—causing researchers to miss insights from would-be respondents with moderate, diversified views (Anderson 1998) and to overestimate associations between-study variables.
Managerial Implications
From a practical standpoint, the results of the analysis should encourage service organizations to strengthen their commitments to further developing an EC. Such commitments offer a series of desirable outcomes for FLEs who represent the organization in their customer-facing roles. By contrast, some notable service brands are learning firsthand about the adverse impact of a questionable corporate culture on their firms’ reputation (e.g., Uber, Wells Fargo; Lublin 2017). What is encouraging, however, is that boards are taking notice and bolstering their oversight of company culture (i.e., culture committees, culture audits) in recognition of the competitive advantages a strong culture offers (Lublin 2017). Such advantages may prove most prominent in competitive markets in which personal relationships play a more important role. Also promising is the ability of an organization’s EC to influence ethical FLE behavior (Schwepker, Ferrell, and Ingram 1997; Weeks and Nantel 1992). In this vein, managers could, for example, implement recruiting and selection efforts, regular ethics training, and strict enforcement of ethical codes (Bass, Barnett, and Brown 1998; O. C. Ferrell, Johnston, and Ferrell 2007).
Managers assume an important role in shaping the organization’s EC as a reflection of their behaviors (Dickson et al. 2001). However, beyond helping create the EC within the organization, the moderator findings from this meta-analytic examination are also instructive to managerial efforts to help the climate take hold under various conditions. For example, understanding that FLEs with less experience have a weaker connection between the organization’s EC and outcomes should lead managers to ramp up their ethical content in FLE onboarding and pay extra attention to new employees to help them grasp ethical prioritizations. Furthermore, managers of FLEs in countries where corruption and collectivism are high must make extra efforts to ensure their EC matriculates to FLE actions, attitudes, and outcomes.
With these implications in mind, this research brings resolution to counterarguments regarding the universality of EC’s positive influence. Some suggest that a demanding EC can create stress and strain for employees. For example, while customer-oriented behaviors aim to generate long-term relational value, this result may come at the expense of short-term sales (Coelho et al. 2010). Combine this potential trade-off with the contingent reward structure that typifies many FLE positions, and the argument that an EC may deplete the cognitive capabilities that normally deflect FLEs’ stressors can follow (Bolander et al. 2017). The results of the analysis, however, show that FLE felt stress is both directly reduced by EC and indirectly reduced through customer-oriented behaviors. That this effect occurs in the frontline context should provide reassurance to service managers that investments in EC and other interrelated areas (e.g., compensation systems, control, and feedback systems) will be met with desirable downstream influences, rather than suppressing certain outcomes.
Limitations and Future Research
This study synthesizes accumulated knowledge germane to EC in the FLE context using a meta-analytic approach. While meta-analysis offers strengths as a methodological technique, this study is not without limitations. The constraints of our inquiry, as well as the shortcomings gleaned from the literature review and coding process, uncover several future research opportunities.
As a processual limitation, the data for the relationships involving our EC construct were generated through our collection and espoused meta-analytical processes. Other relationships necessary to test our SEM, however, came from supplemental management and FLE-related meta-analyses (e.g., Franke and Park 2006; Verbeke, Dietz, and Verwaal 2011; Zablah et al. 2012). Researchers engage in this practice to provide greater parsimony in conducting mediation in meta-analyses in which established meta-analytic relationships (e.g., customer orientation–perceived performance) already exist (e.g., Nicolaides et al. 2014). Nevertheless, the use of meta-analytic relationships from other examinations and populations is a limitation of this study.
Another limitation is the unidirectional nature of the model focusing on downstream effects of EC. As the debate deepens on individual versus organizational drivers of unethical employee acts that may jeopardize customer exchanges, scholars less frequently look upstream to vet the managerial determinants facilitating a stronger EC (Goebel and Weißenberger 2017; Gustafson, Pomirleanu, and John-Mariadoss 2018). Our study focuses on downstream FLE outcomes, given the need for a critical mass of studies in meta-analyses. However, as empirical research in this area continues to amass, antecedent-focused meta-analyses on managerial variables influencing FLE EC may be conducted, such as control systems (Verbeke, Ouwerkerk, and Peelen 1996), ethics training (Valentine 2009), sales quota difficulty (Schwepker and Good 1999), and leadership (Mulki, Jaramillo, and Locander 2009). In addition, variables capturing product (e.g., product complexity/variety; Yi, Dubinsky, and Un Lim 2012), team (e.g., peers’ ethical behavior; Ruiz-Palomino, Martínez-Cañas, and Fontrodona 2013), customer (e.g., customer demandingness; Jaramillo, Mulki, and Boles 2013), or environmental (e.g., market competitiveness; Verbeke, Ouwerkerk, and Peelen 1996) characteristics may be assessed for their causal effect. Furthermore, beyond meta-analytic considerations, our review reveals opportunities for primary data collection on FLE-related antecedents to EC on these same characteristics: product (e.g., product life cycle duration), team (e.g., team cohesiveness), customer (e.g., customer dispersion), and environmental (e.g., market dynamism).
Our review also reveals a preponderance of attention given to main effect models when exploring EC in FLE contexts. While our moderation analysis is restricted in cases in which the bivariate relationship of interest is estimated with a small k, including more moderating measures and sample variables in future research may provide additional insights for the domain. For example, our country-of-origin-based moderators (individualism/collectivism, perceived corruption) suggest that intriguing differences at a national level affect the nature and extent of EC’s impact. Despite this finding, cross-country comparisons in FLE-related EC research are rare (Weeks et al. 2006), while our counter theoretical finding regarding country-level individualism should act as an impetus for similar scholarship to be conducted. Further exploration of the variance that may exist between more established markets and emerging markets may be enlightening. Future research might also account for additional contexts to determine whether EC’s impact on FLE performance is contingent on conditions such as technological turbulence, control systems, and/or customer demandingness.
Empirical insight into EC’s influence on FLEs has been limited to date to cross-sectional inquiry. While myriad important insights have been generated with this type of data, enabling us to show the moderating influence of FLE experience in the meta-analytic model, augmentation in the form of longitudinal data would be welcome. For example, perceptions of EC are not static over time, and assessing the magnitude and nature of temporal variance may be illuminating. For example, the velocity principle suggests that it is not the level of a construct that predicts a result, but rather the change in that construct over time (Palmatier et al. 2013). Researchers could assess whether current levels of EC are driving downstream outcomes or whether the results are more descriptive of increases and decreases in EC perceptions over time.
Finally, researchers have noted that the prediction of FLE behavior and performance often requires insight gathered from multiple levels, such as the organizational and individual levels. To date, however, multilevel-multisource examinations are seldom undertaken to understand the frontline implications of EC, and the current review found a paucity of published articles that relied on such a framework. Future research using multilevel-multisource data would help scholars better understand the variance ascribed to empirical nesting at Level 2 versus Level 1. There are numerous levels and sources of relevance to FLEs that future research could examine, including (1) EC’s impact from both the FLE and the buyer perspectives (Anaza et al. 2015), (2) the hierarchy within and dispersion of the influence of ethics on FLEs operating in teams (Cadogan et al. 2009), (3) the impact of managers’ values on FLEs’ EC perceptions versus individually held attitudes toward ethics and their subsequent climate fit (Fournier et al. 2010; Ingram, LaForge, and Schwepker 2007), and (4) a social network perspective consisting of breadth of sources and depth of levels to extend scholarship on ethics and ethical decision-making within the FLE context (Seevers, Skinner, and Kelley 2007).
Supplemental Material
Supplemental Material, JSR.EthicalClimateMeta.BiographicalParagraph.R5 - Ethical Climate at the Frontline: A Meta-Analytic Evaluation
Supplemental Material, JSR.EthicalClimateMeta.BiographicalParagraph.R5 for Ethical Climate at the Frontline: A Meta-Analytic Evaluation by Scott B. Friend, Fernando Jaramillo and Jeff S. Johnson in Journal of Service Research
Supplemental Material
Supplemental Material, JSR.EthicalClimateMeta.ExecutiveSummary.R5 - Ethical Climate at the Frontline: A Meta-Analytic Evaluation
Supplemental Material, JSR.EthicalClimateMeta.ExecutiveSummary.R5 for Ethical Climate at the Frontline: A Meta-Analytic Evaluation by Scott B. Friend, Fernando Jaramillo and Jeff S. Johnson in Journal of Service Research
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
Note
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.
