When Are Norms Prescriptive? Understanding and Clarifying the Role of Norms in Behavioral Ethics Research

Abstract

Research on ethical norms has grown in recent years, but imprecise language has made it unclear when these norms prescribe “what ought to be” and when they merely describe behaviors or perceptions (“what is”). Studies of ethical norms, moreover, tend not to investigate whether participants were influenced by the prescriptive aspect of the norm; the studies primarily demonstrate, rather, that people will mimic the behaviors or perceptions of others, which provides evidence for the already well-substantiated social proof theory. In this review article, we delineate three streams of norms research in business ethics: behavioral, perceptual, and prescriptive. We argue that by properly categorizing norms, and designing studies to investigate when prescriptions, more than mere mimicry, improve ethical outcomes in organizations, researchers can enhance managers’ efforts to promote ethical outcomes in organizations.

Research on descriptive norm influence has progressed significantly in the past twenty years (for reviews, see Cialdini & Goldstein, 2004; Horne & Mollborn, 2020; Miller & Prentice, 2016; Morris, Hong, Chiu, & Liu, 2015), but recent efforts to extend this research to ethical norms have proved problematic. Behavioral ethics researchers have sought to disseminate norms labeled “ethical,” “moral,” or with quasi-ethical language that implies, without explicitly claiming, that the norms ought to be followed as a matter of ethics. By characterizing descriptive norms as ethical or moral or via quasi-ethical language, organizational ethics research risks committing the naturalistic fallacy (Hume, 1738; Moore, 1903), wrongly blurring the distinction between “what is” (i.e., descriptive norms) and “what ought to be” (i.e., prescriptive norms).

After claiming or implying that the descriptive norms they study are ethical, researchers typically conclude that transmitting these descriptive norms (i.e., influencing people to replicate others’ behavior) promotes ethical decision-making and ethical organizations. These norms, however, often influence others through “social proof,” a form of social influence in which people copy the behaviors of others (Cialdini, Reno, & Kallgren, 1990). Social proof has been well established as a means of transmitting descriptive norms (MacCoun, 2012, 2015). Philosophical analysis, however, suggests that people follow prescriptions differently from how they copy others’ behavior. More specifically, people follow prescriptive norms because of the cognitive “grip” (Korsgaard, 1996: 19) that these norms have on them. This analysis suggests that recent behavioral ethics findings do not show how to promote ethical reasoning but, rather, only how to replicate others’ behaviors.

Furthermore, confusion about the nature of norms can lead to problems not only for researchers but also for practitioners. From a research perspective, imprecise labeling can lead to misleading findings as well as the need to relabel or further parse constructs. For example, behavioral norms researchers may find themselves in the same predicament as organizational behavioral researchers who study “prosocial behavior,” a quasi-ethical labeled construct that rose in popularity in the 1980s (Brief & Motowidlo, 1986; O’Reilly & Chatman, 1986). After a substantial body of research promoted prosocial behavior in work settings, researchers realized that the construct included the types of behaviors that help spread corruption, such as accounting fraud (Mishra, Ghosh, & Sharma, 2022; Warren, 2003). This realization led to the formation of another construct, “unethical prosocial behavior,” intended to capture the unethical acts that can occur while one is behaving prosocially (Umphress & Bingham, 2011). Even though this research is not focused on norms per se, it captures the pitfalls of quasi-ethical labeling.

Understanding what a norm entails is important not only for properly labeling it but also for studying what the norm influences (e.g., reasoning, attitudes, behavior). Without this knowledge, practitioners who apply lessons of descriptive norms with the goal of fostering an ethical organization may find themselves with unintended consequences, as seen with prosocial behavior, or may achieve only short-lived effects. As a further example of the unintended consequences that can be associated with using descriptive norms to foster ethical organizations, consider Cialdini and colleagues’ antilittering studies. Cialdini and colleagues devised several clever experiments that demonstrate how conveying behavioral norms for littering, such as visual demonstrations of littered areas, will not ignite a sense of civic duty and cause study participants to avoid littering (Cialdini et al., 1990; Goldstein & Cialdini, 2007). Rather, such study designs caused higher levels of littering. An initial reaction may be that firms simply need to choose norms wisely, such as those that emphasize a tidy area. In this article, we will assert that even norms that capture an ethical behavior will not improve people’s ethical reasoning or promote ethical organizations that endure over time if they do not evoke moral awareness. Rather, norms that capture an ethical behavior without evoking moral awareness, we will assert, will only achieve short-lived changes in behavior.

To start, we review conceptualizations and differences between descriptive norms (both behavioral and perceptual) and prescriptive norms and explain the foundations of norm confusion in the naturalistic fallacy. Then, we turn to recent behavioral ethics literature on norms and demonstrate how norm confusion can undermine researchers’ results. Finally, we explain how properly distinguishing norms can improve research—making its implications clearer and more impactful—and promote ethical decision-making in organizations.

TYPES OF NORMS

We begin by defining the concepts we will use in our examination, including the concept of norms itself, along with descriptive norms (both behavioral and perceptual) and prescriptive norms. Because our review addresses confusion among some of these concepts, it is important to begin by clarifying how we define them. Our conceptualizations reflect previous distinctions between norms (e.g., Donaldson & Dunfee, 1994; Goldstein & Cialdini, 2007; Morris et al., 2015).

Norms

The concept of norms is used in a variety of contexts to capture the regularity of thought or behavior across a reference group, often with a goal of understanding how such regularities affect individuals within these groups (Horne & Mollborn, 2020; Miller & Prentice, 2016; Morris et al., 2015). Norms may be informally understood and communicated via observation (e.g., basic etiquette, work practices) or formally expressed and codified (e.g., rules, laws) (Miller & Prentice, 2016; Morris et al., 2015). Importantly, norms may diverge, capturing what is expected but not what people actually do or think. We discuss two types of norm content, descriptive and prescriptive, in turn.

Descriptive Norms

Descriptive norms capture the observed or communicated behavioral or thought tendencies across a reference group. They describe how people do behave—or the attitudes that people do have regarding how people should behave—as opposed to how they ought, ethically speaking, to behave. Descriptive norms include both behavioral norms and perceptual norms. We define these norms, explain how they can be transmitted, and discuss examples from the workplace (see Table 1 for illustrations of descriptive norms with quasi-ethical labels).

Table 1: Illustrations of Quasi-ethical Language in Norms Research

Behavioral norms. Behavioral norms capture the observed or communicated behavioral tendencies, or patterns of behavior, of a group (Eriksson, Strimling, & Coultas, 2015; Miller & Prentice, 2016; Morris et al., 2015). Behavioral norms can be transmitted to others via observation or messaging. An employee can adopt a workplace’s behavioral norms regarding meeting attendance, for example, either by observing how coworkers behave or by reading an employee handbook. As Goldstein and Cialdini (2007: 167) explain, people who are influenced by behavioral norms “need not engage in elaborate cognitive processing of the relevant information because applying the heuristic rule ‘I should do what most others do’ is based primarily on the simple observations of others’ situation-specific behaviors.”

In a business context, behavioral norms can relate to hiring, compensation, environmental conservation, lying in negotiations, investment decisions, analysts’ coverage of firms, paying taxes, personal savings, and payments to governmental officials (Bailey, Nofsinger, & O’Neill, 2004; Beshears, Choi, Laibson, Madrian, & Milkman, 2015; Goldstein, Griskevicious, & Cialdini, 2011; Gunia et al., 2012; Hallsworth, List, Metcalfe, & Vlaev, 2017; Kay et al., 2009; Rao, Greve, & Davis, 2001; Schultz et al., 2007; Spicer, Dunfee, & Bailey, 2004; Sunstein, 2016). We call these norms “behavioral” rather than “descriptive” (e.g., Sherif, 1936) to capture the distinction between behavioral and perceptual norms within the descriptive norms family (Cialdini, 2012).

Perceptual norms. Perceptual norms, in turn, refer to the communicated perceptual tendencies of a group. They describe the reference group’s perceptions or attitudes regarding behavior. Perceptual norms can be transmitted via informal or formal messaging. For example, an employee can conform to workplace perceptual norms by engaging in casual conversations with other employees (e.g., in which employees communicate the best way to request vacation time) or reviewing the company’s employee handbook (e.g., a company’s description of how to request vacation days).

Research in this area includes perceptions of how people should be compensated, who should be hired, accounting practices, conservation, consumption beliefs, and the importance of agency theory (Eriksson et al., 2015; Goldstein, Cialdini, & Griskevicious, 2008; Kay et al., 2009; McCarthy & Puffer, 2008; Melnyk, van Herpen, Fischwer, & van Trijp, 2013; Morris et al., 2015; Schultz et al., 2007; Spicer et al., 2004). While some researchers refer to these norms as “injunctive” (Goldstein & Cialdini, 2007) or “normative” (Deutsch & Gerard, 1955; Tepper, Henle, Lambert, Giacalone, & Duffy, 2008), we prefer the term “perceptual” in this context. We call these norms perceptual rather than injunctive or normative because this term signifies perceptions or attitudes regarding appropriate behavior that do not need to be openly stated (i.e., injunctive) or grounded in normative analysis (i.e., normative).

Like behavioral norms research, studies involving perceptual norms tend to examine how individuals respond to norms that are given by the researcher. In their field experiment on delinquent taxpayers, for example, Hallsworth and colleagues (2017: 30) examined how stating “nine out of ten people agree that everyone in the UK should pay their tax on time” affected tax payment. In contrast to the behavioral norm condition (e.g., “Nine out of ten people pay their tax on time”), this experimental condition addresses what people think should happen regarding tax payments. The perceptual norm condition neither describes what people actually do regarding their tax payments nor prescribes why taxpaying is ethical.

Prescriptive Norms

Prescriptive norms, finally, set forth how people ought to behave based on justifications that are independent from either how people actually behave or how they think people should behave. Empirical researchers consider a norm to be prescriptive if it has been morally evaluated (Samland & Waldmann, 2016), if it is morally ideal (Bear & Knobe, 2017), or if the majority of people affected by the norm consider it to be acceptable (Dannals & Miller, 2017). Within philosophy, researchers have justified norms as being prescriptive if they satisfy “certain conditions” that have been specified (Kalantari & Luntley, 2013: 420). Such conditions could include the standards for ethical behavior associated with various ethical theories (e.g., virtue ethics, Kantian ethics, utilitarianism, justice, contractualism) (Windsor, 2004) as well as fundamental principles, such as the Universal Declaration of Human Rights (Carasco & Singh, 2008; Donaldson, 1989; Donaldson & Dunfee, 1994; Frederick, 1991; Hartman, Shaw, & Stevenson, 2003; Jackson, 2000; Rowan, 1997; Scherer, 2015). Of these standards, those associated with ethical theories dominate the business ethics literature (Audi, 2010; Derry & Green, 1989).

Prescriptive norms can be transmitted via messaging combined with people’s recognition that particular norms are prescriptive for them (and other members of the relevant reference group) based on their role or position in society (Spicer et al., 2004). For example, employees can learn their workplace’s prescriptive norms by reading a corporate ethics code (e.g., stating that the company values confidentiality) and recognizing that the prescriptive norms pertain to them as employees. Similarly, professions often set forth prescriptive norms for members of the profession in professional codes of conduct (Jamal & Bowie, 1995; Lückerath-Rovers & De Bos, 2011). Research in prescriptive norms considers whether norms related to exposing workers to hazardous materials, reneging on contracts, paying a living wage, avoiding market failure, and discriminatory employment policies, among others, meet prescriptive standards (Bailey & Spicer, 2007; Donaldson & Dunfee, 1994; Spicer et al., 2004; Van Oosterhout, Heugens, & Kaptein, 2006). While some researchers refer to these norms as “normative” (e.g., Frederick, 1991) or “evaluative” (e.g., Dancy, 1995), we prefer the term “prescriptive” in this context. Although prescriptive norms are grounded in normative analysis (i.e., are normative), we do not use this term to set forth our claims about norm confusion because “normative” has been used ambiguously in the relevant literature to signify what we call “perceptual” norms (e.g., Deutsch & Gerard, 1955; Tepper et al., 2008). The concept of “evaluative,” in turn, fails to capture the way in which prescriptive norms require people, ethically speaking, to manifest certain behaviors (Kolodny, 2005).

Behavioral and perceptual norms can be prescriptive when they are justified by standards that are independent of people’s observations of, or perceptions concerning, the reference group’s behavior. Labeling behavioral or perceptual norms moral or ethical, however, does not make them so. When researchers call a behavioral or perceptual norm moral or ethical without justifying why the reference group implicated by the norm ought to follow this norm, the attribution of a prescriptive status to the norm is unjustified. For the remainder of the article, we take the terms prescriptive norm, ethical norm, and moral norm to be interchangeable.

FOUNDATIONS OF NORM CONFUSION

We discuss the conceptual underpinning of norm confusion in the distinction between the descriptive realm (“what is”) and the prescriptive realm (“what ought to be”). Then, we explain how behavioral ethicists’ study of descriptive norms in business, alongside normative ethicists’ research on prescriptive norms, sets the stage for norm confusion in the business ethics literature.

The Descriptive–Prescriptive Distinction

Philosophers have distinguished between the descriptive realm, “what is,” of empirical phenomena in the world and a prescriptive realm concerning “what ought to be,” whose standards people can strive to meet but that does not necessarily exist presently (Hume, 1738; Moore, 1903). Scholars disagree about whether this distinction is clear (e.g., Donaldson, 1994) or fuzzy (e.g., Hartman, 2008; Putnam, 2002; Rosenthal & Buchholz, 2000). We discuss demonstrations that the distinction is clear, along with several reasons to think the distinction is fuzzy, in turn. To support our position that a distinction exists between “what is” and “what ought to be,” we end by explaining why we are not convinced by the discussion of reasons to think the distinction is fuzzy.

Reasons to Think the Distinction Is Clear

The distinction between “what is” and “what ought to be” is clear, in some scholars’ views, to the extent that confusing these concepts represents a logical fallacy. That is to say, the descriptive–prescriptive distinction is founded in the so-called naturalistic fallacy (Goodpaster, 1985; Hume, 1738; Moore, 1903; Victor & Stephens, 1994), which refers to people’s unjustified tendency to assume that the world as it presently exists is how the world ought, ethically speaking, to be. When large numbers of people behave in the same way or approve of the same behaviors, for example, people tend to assume that such behaviors are ethical. This assumption is fallacious because people can behave in various ways in most situations, some of which meet ethical standards and others of which do not. Both the number of people behaving in a particular manner and people’s dis/approval of the behavior are independent of whether the behavior is ethical.

There is empirical support for the view that people tend to assume, without justification, that common, or commonly approved, behaviors are ethical. Eriksson and colleagues (2015), for example, designed experiments involving fictitious behaviors (e.g., “phooshing” and “quining”) to demonstrate how easily people are swayed by behavioral and perceptual norms. Study participants were told that phooshing and quining were either commonly exhibited behaviors (i.e., behavioral norms in our categorization) or socially desirable behaviors (i.e., perceptual norms in our categorization). From these descriptions, the study participants inferred that the fictitious behaviors were morally appropriate (i.e., prescriptive norms in our categorization). It is obvious that “phooshing” and “quining” are not prescriptive behaviors, however, because these terms do not signify any behaviors at all. In pointing out participants’ errors in assuming that fictitious behaviors are ethical, Eriksson and colleagues’ (2015) study thus documents that researchers can clearly distinguish “what ought to be” from “what is” (even if participants did not).

Reasons to Think the Distinction Is Fuzzy

Scholars have also challenged the is–ought distinction; we discuss two reasons to think that this distinction is fuzzy rather than clear. First, scholars have argued that descriptions of the world are inextricably intertwined with the social contexts from which they arose, which include assumptions about the way the world ought to be (e.g., Harris & Freeman, 2008; Putnam, 2002). Thus social scientists’ empirical observations always presuppose certain normative claims (e.g., Sen, 1987). One example of this phenomenon is a team of behavioral ethicists perceiving that a behavioral norm exists and inferring that following this norm is optimal, ethically speaking. For example, Goldstein and colleagues (2008) study how most effectively to influence hotel guests to reuse towels. These researchers need not assume that hotels ought, ethically speaking, to influence guests to reuse towels to conduct this research, of course. If hotels influence guests to reuse towels to create a pretext for cutting the housekeeping staff, for example, then these behavioral ethicists might regard the behavior as wrong. The idea that social scientists cannot help but regard the empirical phenomena they study as (either) right or wrong, however, suggests that the distinction between “what is” and “what ought to be” is fuzzy, not clear.

A second challenge to the descriptive–prescriptive distinction relates to “thick ethical concepts” (Williams, 1985), such as cruel or courageous, that combine descriptive and prescriptive elements. Whereas these concepts ostensibly describe behaviors in the world, such as “the military’s killing of civilians was cruel,” the descriptions have prescriptive significance to the extent that calling behavior “cruel” implies that one ought not to act that way (cf. Putnam, 2002). By making a prescriptive conclusion (e.g., one should not act cruelly) arise from a description of the world (e.g., acting cruelly causes pain deliberately), thick ethical concepts appear to blur the is–ought distinction (Väyrynen, 2021).

Our Position

We advocate distinguishing between descriptive and prescriptive norms more clearly because we believe it matters for promoting ethical reasoning and ethical organizations. As such, we endorse this distinction at least to a certain extent. To support our position, we respond to the challenges offered in the previous sections. First, although we agree that social scientists naturally take positive or negative attitudes regarding the phenomena they study, we deny that they do so inescapably. We take it as obvious, rather, that social scientists can take a step back and critically examine their hunches concerning the rightness or wrongness of various behaviors. As empirical support for this suggestion, we note that people’s attitudes can change over time and vary across geographical locales. In their study of increasing tax compliance via exposing people to behavioral norms regarding others’ tax compliance, for example, Hallsworth and colleagues (2017) assume that it is ethically obligatory to pay taxes on time. Although this view arguably holds for most people in the contemporary United States, it is not objective or timelessly true. Societies in other times and other places have enjoyed paradigms of tax avoidance (e.g., Ravenda, Argilés-Bosch, & Valencia-Silva, 2015).

Regarding thick ethical concepts, in turn, we agree that such concepts exist. Indeed, what we call “quasi-ethical” behavioral norms could be construed as thick ethical concepts. We deny, though, that they reveal “what ought to be” rather than “what is.” Consider the example of honesty (e.g., Blay, Gooden, Mellon, & Stevens, 2019; Gino, Ayal, & Ariely, 2009) or responsibility (e.g., Blay et al., 2019). Describing behavior as honest or responsible implies that the behavior is ethical. In the same way that cruel behavior can be ethical, such as when a general bombs a building with civilians inside in order to blow up a munitions factory next door (as discussed in Additional Protocol I to the Geneva Conventions) (Hurka, 2005), honest or responsible behavior can be unethical. Honest behavior can be unethical, for example, when a manager’s family member asks her to reveal a confidential aspect of her company’s business strategy so that the family member can decide whether to purchase shares of the company’s stock. Similarly, the behavior of low-level employees who dutifully follow managers’ admonitions to continue selling a dangerous product would typically be considered unethical. This is because thick ethical concepts only imply that people ought to behave in certain ways. Classifying a behavioral norm as a thick ethical concept does no more to make it prescriptive than calling the norm ethical or moral. Thus the is–ought distinction holds, despite these challenges.

Norms in Behavioral and Normative Ethics

Next, we explain how behavioral ethicists’ study of descriptive norms in business alongside normative ethicists’ research on prescriptive norms in business set the stage for the norm confusion we address. We survey research concerning both descriptive norms and prescriptive norms, along with accounts of the relationship between the descriptive and prescriptive realms in the business ethics literature. Then, we explain the contribution we seek to make regarding the relationship between the descriptive and prescriptive realms in the business ethics literature.

Research Concerning Descriptive Norms

Study of descriptive norms began with Sherif’s (1936) research on social proof. The initial studies took place in a laboratory setting in which study participants were asked how much a beam of light moved. Sherif found that people adjusted their estimate of the light’s movement to reflect the group’s estimate. Even when asked privately and away from the group, people’s estimates aligned with the group’s estimate and sometimes strengthened. This led researchers to theorize that individuals rely on the behavior of others (as “social proof”) to determine how to behave, especially in uncertain situations.

The most influential body of research supporting social proof in a business context stems from the work of Cialdini and colleagues (Goldstein & Cialdini, 2007; Goldstein et al., 2011; Schultz et al., 2007). In a variety of settings related to water and energy use, Cialdini and colleagues demonstrated how messages that describe common behaviors affect people’s behavior. For example, Goldstein and colleagues (2008: 476–77) promoted the reuse of towels in hotels by communicating behavioral norms through messages like “in a study conducted in Fall 2003, 76% of the women and 74% of the men participated in our new resource savings program by using their towels more than once.”

Research Concerning Prescriptive Norms

At the same time that empirical scholars were extending social proof theory and applying it in a business setting, normative scholars were investigating the use of ethical theories and prescriptive norms in business decision-making. This literature addresses such topics as people’s ethical obligations in the workplace (e.g., Donaldson & Dunfee, 1994), under what conditions people follow prescriptive norms (as opposed to the local community’s behavioral norms) in the workplace (Spicer et al., 2004), how to generate substantive (prescriptive) norms for ethical decision-making in business organizations (Van Oosterhout et al., 2006), and how ethical theories can be used to build international business norms (Windsor, 2004). Business ethics research concerning prescriptive norms typically involves which norms are prescriptive in a business setting and which ethical standards should be used to determine which norms are prescriptive in a business setting.

Research Concerning the Is–Ought Distinction in Business Ethics

Several articles have examined the is–ought distinction and the role that it can play in empirical and normative research. The jumping-off point for these articles is typically the authors’ confoundment that empirical and normative scholars profess to investigate the same problem in business (namely, unethical behavior) while lacking appreciation, or even understanding, of one another’s distinctive methodologies and aims (e.g., Treviño & Weaver, 1994). De Los Reyes, Kim, and Weaver (2017), for example, discuss differences between the domains and difficulties in reconciling empirical and normative research, while defending the importance of at least including both perspectives in business ethics courses. Treviño and Weaver (1994) set forth differences between how empirical and philosophical scholars address business ethics problems, with an aim of facilitating their continued exploration of and collaboration regarding the business practices that both regard as problematic, ethically speaking. Weaver and Treviño (1994), in turn, set forth various positions that this distinction can play in business ethics research: a strict role, such that these two faces of business ethics have little to contribute to one another; a flexible role, such that empiricists and philosophers can communicate with one another about their business ethics research to their mutual benefit; or a rejection of the distinction, such that scholars understand human behaviors and perceptions as unavoidably expressing ethical concerns. Donaldson and Dunfee (1994), finally, have offered a normative approach, integrative social contracts theory, that includes empirical information, thus creating at least one decision-making framework that embodies both approaches.

Our Contribution

Our efforts to fruitfully combine prescriptive and descriptive research perhaps most resemble the “meaningful dialogue” that Singer (1998: 482) attempts to construct between these disciplines. Unlike many earlier efforts to investigate and offer strategies for bridging the descriptive–prescriptive divide, we do not aim, primarily, to analyze differences between empirical and normative research or to propose strategies for reconciling them. Rather, we identify a particular problem in recent empirical research concerning norm influence (researchers’ tendency to assume that transmitting descriptive norms can have a prescriptive effect) and seek to redirect the research toward a strategy that is better suited to promoting ethical organizations. As we discuss in the next section, several recent articles have built on social proof demonstrations concerning how to produce certain behaviors in populations of people, adding ethical assumptions about the norms they aim to disseminate. These articles have thus treated prescriptive norms as though they are no different from behavioral or perceptual norms. Our contribution to the business ethics literature concerning the relationship between prescriptive and descriptive research, then, is to analyze this tendency in behavioral ethics research, show why it is harmful, and propose solutions.

Although our aim is thus distinct from previous discussions of the is–ought distinction in business ethics, these discussions both shape and shed light on our analysis. For example, we note that our approach falls somewhere between Weaver and Treviño’s (1994) flexible and integrated alternatives. We clearly hold that the is–ought distinction has a place in business ethics; in our efforts to support empirical researchers gaining insight into prescriptive norms, though, we acknowledge that the disciplines are integrated enough to permit empirical researchers to investigate prescriptive norms. Whereas we agree with Victor and Stephens (1994) that business ethics is a “single unified field,” we disagree that conceiving of business ethics as a unified field obviates the is–ought distinction. Prescriptive norms represent the world as it ought to be, rather than as it is; empirical researchers can nonetheless study how people cognize these norms (rationally endorsing, and holding themselves to be bound to follow, particular prescriptive norms, independently of whether people generally behave in line with, or approve of, the norms) and how such cognitions affect people’s behavior.

NORM CONFUSION IN RECENT BEHAVIORAL ETHICS RESEARCH

In this section, we document some of the problems that failing to distinguish between descriptive (behavioral or perceptual) and prescriptive norms has caused in recent behavioral ethics research regarding norm influence. We focus on studies in which researchers have labeled descriptive norms “ethical” (e.g., Chen et al., 2017; Flynn & Wiltermuth, 2010; Gunia, Wang, Huang, Wang, & Murnighan, 2012) or “moral” (e.g., Blay et al., 2019; Eriksson et al., 2015) or have used quasi-ethical language that implies, without explicitly claiming, that the norms ought to be followed as a matter of ethics (e.g., Alcott & Mullainathan, 2010; Amini, Ekstrom, Ellingsen, Johannesson, & Stromsten, 2017; Andersson & Pearson, 1999; Bhanot, 2021; Blay et al., 2019; Briscoe, Chin, & Hambrick, 2014; Chang, Milkman, Chugh, & Akinola, 2019; Gino et al., 2009; Goldstein et al., 2008; Goldstein et al., 2011; Hofenk, van Birgelen, Bloemer, & Semeijn, 2019; Miller & Prentice, 2016; Molenmaker, de Kwaadsteniet, & van Dijk, 2014; Mun & Jung, 2018; Roos, Gelfand, Nau, & Lun, 2015).

We discuss recent studies in terms of whether the study identifies behavior as prescriptive and whether the study examines how the prescriptive status of the behavior affects the study’s outcome. To elaborate them, we walk through several recent studies in which researchers confuse descriptive norms (behavioral or perceptual) with prescriptive norms across a range of situations, including conserving natural resources, discouraging fare evasion, and promoting truth telling.

Conserving Natural Resources

Bhanot (2021), for example, examines how perceptual norms affect conservation by measuring how water bills that compare consumers’ personal water use to community norms and include a smiley, neutral, or frowny face affect future water use. The messaging conveys a perceptual norm by indicating which levels of water use, relative to community norms, gain a “social judgment” of approval or disapproval (Bhanot, 2021: 30). The author confuses perceptual and prescriptive norms to the extent that he implies that consumers ought, ethically speaking, to conserve water. He makes this implication via quasi-ethical language about “what others consider ‘good’ and ‘bad’ behavior” (Bhanot, 2021: 30).

Our first question is whether the study has identified behavior as prescriptive. In this case, the study makes some progress in identifying behavior as prescriptive. Although Bhanot (2021) does not explicitly prescribe that people ought to conserve water, his quasi-ethical language does imply that consumers should use less water than others in their community on ethical grounds. Bhanot’s implied prescription is not fully articulated, and remains vague, however, to the extent that consumers are not instructed to modify their behaviors with respect to water use in specific ways. Rather, they receive a smiley or frowny face based exclusively on how their water use compares with other people’s (with whose water use the consumer receiving a smiley or frowny face is not personally familiar). Thus Bhanot does not clearly prescribe behavior to water consumers.

Our second question is whether the study examines how the prescriptive status of the behavior affects the study’s outcome. Bhanot (2021) does not consider how the (implied) prescriptive status of water conservation influences whether consumers increase or reduce their water use. In his field study, water consumers received messaging in their water bills consisting of a smiley, neutral, or frowny face depending on how the consumers’ water use compared to that of others. Consumers received the messaging in water bills over several billing periods; the messaging changed from bill to bill as consumers modified their water use. Bhanot did not study the reasons why consumers modified their water use, however, such as preferring to receive a smiley face rather than a frowny face for aesthetic reasons, preferring to receive a smiley face rather than a frowny face because of their estimation for social approval, or believing that it is ethical to conserve water and striving to bring their behavior into line with an ethical prescription.

Discouraging Fare Evasion

Ayal et al. (2021), in turn, use messages conveying behavioral norms, combined with an image of watching eyes, to influence people’s behavior with respect to fare evasion at a train station. They study norms such as “in this station, 90% of all individuals purchase and validate their ticket” (12), which are behavioral to the extent that they describe what people do. The authors confuse behavioral and prescriptive norms in their assumption that influencing people to follow a descriptive norm can have an ethical outcome, including helping to preserve trust and discourage “negative social norms” (9).

Our first question is whether the study has identified behavior as prescriptive. In this case, Ayal and colleagues (2021) have identified behavior as prescriptive. They study specific behaviors: purchasing and validating train tickets or failing to purchase and validate train tickets. To indicate that the former behavior is prescribed and the latter is not, Ayal and colleagues describe failing to purchase and validate train tickets as “unlawful,” thus invoking the law as a prescriptive standard.

Our second question is whether the study examines how the prescriptive status of the behavior affects the study’s outcome. In this case, the messaging concerning other passengers’ behavior, along with an image of watching eyes, led to the study’s outcome in reducing fare evasion. The fact that participants began to purchase and validate tickets after viewing Ayal and colleagues’ (2021) posters does not indicate either that they now consider not purchasing and validating tickets to be unethical or that they have suddenly become honest and/or ethical in purchasing and validating tickets. Rather, the study informed them that an alternative behavioral norm, purchasing and validating tickets, was salient, and they began to follow that behavioral norm. They determined their behavior in both cases, arguably, based on how they understood other train travelers to behave. As such, Ayal and colleagues’ results in reducing fare evasion do not provide evidence about motivating people to follow prescriptive norms.

Promoting Truth Telling

In their study of conversations about truth telling, finally, Gunia and colleagues (2012: 31) observe that participants tend to tell the truth after conversation partners have informed them that “most people would b[e] honest in this … situation.” This norm is behavioral to the extent that it describes what people do (be honest). It conflates descriptive norms with prescriptive norms to the extent that the authors imply that being honest is ethical.

Our first question is whether the study has identified behavior as prescriptive. In this case, Gunia and colleagues (2012) have identified prescriptive behavior. Study participants have two options, receiving messaging that implies that either one of the behaviors is ethical (telling the truth) and the other is not (doing what will maximize the amount of money the participant receives in the study). Although the norm is couched in the language of behavioral norms, it also prods participants to decide for themselves which option constitutes honest (a quasi-ethical term) behavior and which does not.

Our second question is whether the study examines how the prescriptive status of the behavior affects the study’s outcome. In this case, labeling the truth telling (behavioral) norm moral leads researchers to conclude that ethically oriented conversation produces ethical behavior (Gunia et al., 2012: 19). The researchers do not, though, study how the prescriptive status of truth telling influences participants’ behavior. Truth telling could be the result of social proof, rather than the ethical outcome of having adopted the prescription that people should tell the truth.

HOW NORM CONFUSION HARMS RESEARCH AND PRACTICE

The norm confusion documented in the previous section has two problematic downsides for research and practice. First, the studies fail to raise moral awareness. This entails that participants could simply shift behaviors when a novel behavioral norm is introduced, whether or not the novel behavioral norm is ethical. In other words, the studies offer no reason to think that participants will retain the behavior, much less the proclivity to behave ethically, in alternative contexts. Second, the studies engage in manipulation. We discuss these problems in turn.

Failing to Raise Moral Awareness

The studies highlighted in the previous section are particularly instructive because, as discussed in that section, some of the behavioral norms associated with the studies are prescriptive. Even prescription-satisfying behavioral norms, though, may not help to foster enduring ethical outcomes in organizations if people within the organizations are not aware that the norms are ethical. Transmitting behavioral norms (e.g., via conversations about the norm or messaging about the norm combined with images of watching eyes) neither conveys appreciation that the behavior is ethical, raises the moral awareness of those to whom the norm is transmitted, nor otherwise engages the ethical reasoning of those to whom the norm is transmitted. As such, recent empirical studies of norm influence offer no reason to think that the participants to whom they transmit (prescription-satisfying) behavioral norms will retain the ethical behavior. We begin by analyzing these studies. Then, we discuss philosophical theory that supports our analysis and respond to psychological theory that challenges our analysis.

Truth Telling

Gunia and colleagues (2012), for example, do not consider how the ethical status of the truth telling (behavioral) norm affects whether people follow the norm. Study participants might have been influenced to tell the truth in a social proof manner, simply mimicking what they understood to be the relevant behavioral norm. In this form of influence, participants would lie after having been informed that most people lie, behaving in an apparently unethical manner after having been exposed to an apparently unethical descriptive norm. If they were influenced by the ethical status of the truth telling behavioral norm, by contrast, then the prescriptive norm may take grip, and people’s behavior might meet the standard of the prescription (telling the truth) regardless of the descriptive norms to which they were exposed. Participants might behave in an ethical manner even after having been exposed to a norm about unethical behavior, for example, if the ethical status of the (unethical) descriptive norm triggers them to think about the ethical status of their own behaviors and to strive to meet the ethical standards (i.e., prescriptive norms) that they recognize.

Paying Fares

Although Ayal and colleagues (2021: 9) both identify the norm that they study, purchasing and validating tickets at a train station, as quasi-ethical (“lawful”) and acknowledge their research’s grounding in social proof theory, their research remains emblematic of the problem that we highlight. The fact that participants began to purchase and validate tickets after viewing Ayal and colleagues’ posters entails neither that they are aware that not purchasing or validating tickets raises an ethical issue nor that they appreciate that their behavior is now ethical; rather, the study informed participants that an alternative behavioral norm, purchasing and validating tickets, was salient, and they began to follow that behavioral norm. Although purchasing and validating tickets appears to conform to a prescriptive norm, Ayal and colleagues’ results do not show that participants will behave in a way that meets ethical standards in any other circumstances.

Conserving Natural Resources

Bhanot (2021) further illustrates our worry that behavioral norms can influence people’s behavior but that being influenced by behavioral norms might not dispose people to behave more ethically in situations not governed by the specific behavioral norm. Like Gunia and colleagues (2012) and Ayal and colleagues (2021), Bhanot (2021) does not demonstrate how to make the shifts in participants’ water usages grip participants, regardless of shifts in the descriptive norms (and smiley or frowny faces) to which they are exposed. When the next behavioral norm is presented, the consumers in Bhanot’s study could simply shift their water usage back to their old habits.

Another recent study illustrates this worry well. In a study of nudges, Wu and Paluck (2021) found that images of golden coins, which convey good luck in Chinese culture, could influence Chinese workers’ behavior. More specifically, placing images of golden coins on the floors of a Chinese workplace discouraged workers from littering there. When the coins were removed without explanation, however, workers resumed their habit of leaving garbage on the floor, leading Wu and Paluck to conclude that “golden coins were no longer effective when they were removed, meaning that the effect was bound to the observation of the physical coins” (49). When researchers neither conveyed to study participants that the behavioral norm of not littering is ethical nor engaged their moral awareness or ethical reasoning about littering in other ways, the messaging associated with the gold coins did not take grip.

Theoretical Support

These problems suggest that norm influence may be unable to foster long-lasting ethical behavior in organizations across a range of circumstances. Because the descriptive norm studies neither raised study participants’ moral awareness nor persuaded them to change their behavior by reasoned arguments, it is questionable that their ethical reasoning shifted. Absent a shift in ethical reasoning, though, there is no reason to think that study participants will behave more ethically in contexts outside the specific scenarios involved in the research. This is because understanding why a behavior is ethical allows an employee to grasp when the situation has changed. Without an understanding of what makes the behavior ethical, an employee’s behavior will be ethical only so long as everything stays exactly the same. According to this concern, then, norm influence will not promote ethical reasoning even when the descriptive norms that individuals mimic do meet ethical standards.

We discuss theoretical research from philosophy that supports this analysis and respond to psychological research that challenges it. Regarding philosophy, most philosophical accounts of ethical action require ethical actions to be intentional (e.g., Alzola, 2015; Frankfurt, 1971; Korsgaard, 1996; Mele & Moser, 1994; Setiya, 2008). The studies we have discussed, though, do not examine what people receive and absorb when they are influenced by descriptive norms. The studies of norm influence we have discussed seek neither to persuade people via reasoned arguments nor to engage them as rational beings. Rather, the studies try to influence people’s behavior by subjecting them to social pressure. The behaviors documented by behavioral ethics researchers can appear, in this sense, to be acts of conformity with descriptive norms without engaging in ethical reasoning. To promote ethical organizations, philosophical accounts of ethical action suggest, by contrast, that researchers need to learn if (and, if so, how) the ethical status of descriptive norms to which participants are exposed influences what participants decide to do.

Against this viewpoint, it could be argued that behavioral or attitudinal conformity with a norm that meets an ethical standard, as appears to occur in Gunia and colleagues (2012), Ayal and colleagues (2021), Bhanot (2021), and Wu and Paluck (2021), is (at least) a step in the right direction in terms of securing people’s conformity with socially desirable aims. An important account of moral behavior within the business ethics literature, in fact, holds that moral behavior is more a matter of intuition and social conditioning than of abstract moral thought (for a review, see Weaver, Reynolds, & Brown, 2014). According to this account of moral behavior, inculcating (prescription-satisfying) behavioral norms in study participants could help to foster ethical behavior in organizations, even if the resulting behavior only (unintentionally) conforms with prescriptive standards, rather than being intentionally ethical.

We agree that behavioral conformity with a norm that meets an ethical standard is a step in the right direction. So long as businesspeople confront various decision-making scenarios, in divergent circumstances and across time, however, we hold that this strategy of producing ethical behavior falls short vis-à-vis promoting ethical organizations. As Stansbury and Barry (2007: 253) have theorized with respect to corporate ethics programs, forcing people to behave in an apparently ethical manner, rather than “enabling” them to reason what is the ethical way to behave in unfamiliar situations, deprives them of the skill sets they need to address complex situations effectively and achieve ethical outcomes over time. As such, we infer that recent behavioral ethics studies that transmit prescription-satisfying behavioral norms, without engaging moral awareness, will have limited value in fostering ethical behavior in organizations.

Even theorists who hold the alternative (intuition-based, nonintentional) view of moral action, moreover, have reason to study prescriptive norms. On our interpretation, psychological accounts of automatic, noninferential moral judgments (e.g., Reynolds, 2006, 2008) suggest that prescriptive norms could grip people more deeply and durably than descriptive norms. Consider research on “moral awareness” (Jones, 1991; Rest, 1986; Reynolds, 2006, 2008), for example, which entails a person’s understanding that “a situation contains moral content” and should be “considered from a moral point of view” (Reynolds, 2006: 233). Reynolds claims that descriptive norms alone are not able to raise moral awareness. He holds that “something must occur that makes the norm vivid to the decision maker and salient to a moral context” (Reynolds, 2006: 234). On Reynolds’s (2008: 1028) view, that “something” could be a “chronically accessible framework of morality,” such as moral attentiveness. On our view, the “something” could be the norm’s prescriptive status. In making claims about what people ought to do, rather than what they actually do, prescriptive norms provide a framework of morality that triggers moral awareness.

Whereas moral attentiveness could lead to “both specific moral behaviors that occur at a single point in time and more abstract or global behaviors, such as general moral conduct, that represent the sum of innumerable moment-to-moment behaviors” (Reynolds, 2008: 1029), transmitting a descriptive norm without the cognitive framework of moral attentiveness would not have this salubrious effect. The cognitive frameworks implicated in moral awareness cause individuals to acknowledge and appreciate the moral dimensions of new stimuli; we also expect prescriptive norms to extend to new stimuli, beyond the specific situation in which the prescriptive norm was introduced.

Manipulating People in Service of Attractive Ends

Along with this issue, a concern about manipulation arises in studies that influence individuals to behave in ways in which they did not choose. Norm confusion abets this problem to the extent that researchers may feel justified in manipulating people, an otherwise wrongful behavior, so long as researchers are manipulating people to behave in ways that they consider to be ethical. Several business ethics scholars criticize this form of manipulation as unethical.

In addressing the ethics of manipulative advertising, for example, Beauchamp and colleagues (1984) argue that manipulation is unethical because it impermissibly interferes with people’s autonomy to make up their minds for themselves about what they wish to do. In Gunia and colleagues’ (2012), Ayal and colleagues’ (2021), Bhanot’s (2021), and Wu and Paluck’s (2021) studies, participants do not make up their minds to tell the truth, pay their train fares, reduce water usage, or clean up after themselves at work, respectively. Rather, researchers manipulate them to behave in these manners by subjecting them to the influence of a salient behavioral norm.

Wilkinson (2013), in turn, argues that manipulation—specifically focusing on manipulative nudges—violates autonomy. Although manipulative nudges do not change the alternatives for decision-making that a person faces, they corrupt the process by which people go about deciding between those options (Raz, 1986; Wilkinson, 2013). The process by which people decide whether to tell the truth, reduce their water usage, or clean up after themselves does appear to have been influenced by the studies’ frameworks, at least to the extent that participants do not decide to behave in these manners because they consider such behaviors to be ethical. Rather, in Gunia and colleagues’ (2012), Ayal and colleagues’ (2021), Bhanot’s (2021), and Wu and Paluck’s (2021) studies, participants follow a behavioral norm, the salience of which researchers have highlighted.

FUTURE RESEARCH: REDIRECTING NORM RESEARCH

In Figure 1, we map the current descriptive norms research with the goal of reorienting it toward prescriptive norms research to better promote ethical organizations. We propose that behavioral ethics researchers begin by determining if the norm transmits a prescription (e.g., people ought to tell the truth). If it does not, then we suggest the researchers consider if an implicit prescription is evoked. For instance, descriptive norms labeled with thick concepts, such as benevolence, honesty, or responsibility, could have embedded prescriptions (e.g., people ought to be benevolent). If the norm transmits a prescription or an implicit prescription, then the researcher should consider if the study design gauges whether the prescription itself influences the ethical outcome. If the study design isolates the prescription as the source of influence, then the study constitutes prescriptive norms research. If the study does not gauge the effect of the prescription, then the research may simply constitute social proof research.

Figure 1: Mapping Descriptive Norms Research and Reorienting toward Prescriptive Norms Research

We acknowledge that there are two forms of social proof, which pose different levels of concerns. For the nefarious social proof research, we assert that research that uses quasi-ethical labels but does not test the transmission of a prescriptive norm could be used to spread unethical behavior. Much like the original prosocial behavior literature, which inadvertently promoted behaviors that facilitated collective corruption, research on norms with quasi-ethical labels could be used for unethical ends, despite the quasi-ethical label. For the benign social proof research, we do not have the same concerns regarding unintended unethical behavior. The research, however, does little to explain how managers can promote ethical organizations because the norm does not transmit a prescription. Without transmitting a prescription, the norm may not raise moral awareness and subsequently cause shifts in employees’ ethical reasoning. Thus, although benign social proof research may cause ethical behavior, the results could be temporary. Even though we view nefarious and benign social proof research as posing different risks to future research, we suggest that neither type of research will provide meaningful findings about how to promote ethical organizations unless the study designs are revised. Here we lay out a pathway for future research that more deeply probes what is transmitted by descriptive norms and encourages future norms research that incorporates the transmission of a prescription and gauges the prescription’s effects.

The Extent of Descriptive Norm Influence

To understand the influence of descriptive norms, we recommend returning to previous research but extending the findings to different contexts and gauging what is transmitted through the norm. For instance, Gunia and colleagues’ (2012) research provides an excellent foundation for examining many questions related to the truth telling norm, which could be used in a variety of contexts to understand not only its influence but the degree to which individuals understand the prescriptive aspects of the norm. For example, the effectiveness of the truth telling norm could be studied in contexts in which telling the truth conflicts with legal or role obligations (e.g., a scenario in which a construction firm is asked for a project bid by a competitor who seeks to engage in collusion). To expand on the tension between the truth telling norm and obligations, study participants could be presented with the truth telling norm and told they must decide whether to disclose a harmful product defect to a customer, even though the information is protected by confidentiality rules. These scenarios would provide important information regarding the nature and limits of a truth telling norm. If study participants follow the truth telling norm across scenarios, whenever it leads to collusion or violates confidentiality rules, then researchers will have robust evidence of social proof theory and must search for organizational interventions that curtail the negative effects of descriptive norms. If the study participants only follow the truth telling norm when the law supports telling the truth or there is a threat of harm, then the findings suggest that ethical limits exist for the truth telling norm; people do not follow the norm across contexts. Rather, individuals intentionally tell the truth when doing so does not violate legal obligations or prevents egregious harm. Such findings would suggest that the study has transmitted the prescriptive aspect of the truth telling norm. As an important next step, researchers should ask study participants for explanations of their logic or reasoning to probe what is transmitted.

Broadening the application of descriptive norms to new contexts could test the limits of conservation research too. What happens, for example, when the perceptual norms communicated through frowny and smiley faces are used to orient customers toward wasteful behavior? If customers are told that most people believe that climate change is not real, recycling requires more effort than it is worth, or most people do not recycle, will customers stop recycling? Here again, researchers will have an opportunity to determine if the messaging related to conservation is simply an extension of social proof or if ethical reasoning limits the boundaries of the norm transmission such that people will not follow a new descriptive norm if it violates certain ethical standards. These study designs could be applied to other areas where organizations struggle to steer employees toward ethical standards, such as those areas tackled in corporate codes of conduct (e.g., gift giving, conflicts of interest, sexual harassment, discrimination) (Carasco & Singh, 2008; Donaldson, 1985, 1989; Frederick, 1991; Hsieh, 2006; Kaptein, 2011; Kaptein & Schwartz, 2008; Treviño, Butterfield, & McCabe, 1998).

If norms related to truth telling and conservation behaviors are so effective that they can easily reorient individuals toward unethical or ethical behavior, then managers need to ensure that descriptive norms align with prescriptive norms in their workplaces. In the remainder of this section, we discuss how to study the gap between descriptive and prescriptive norms.

Studying Descriptive and Prescriptive Norms

We believe that the most promising and important research on ethical organizations lies at the intersection of descriptive and prescriptive norms research. More research needs to be aimed at addressing situations when descriptive and prescriptive norms do not align.

To capture prescriptive norms, researchers could leverage the existing research on organizational, professional, or international codes of conduct (Carasco & Singh, 2008; Donaldson, 1985, 1989; Frederick, 1991; Hsieh, 2006; Kaptein, 2011; Kaptein & Schwartz, 2008; Treviño et al., 1998) and understand how exposure to the prescription affects ethical reasoning. In Table 2, we demonstrate how prescriptive norms that appear within the business ethics literature can conform to or depart from descriptive norms (behavioral and perceptual) and serve as a starting point for empirical studies.

Table 2: Illustrations of Convergence and Divergence of Descriptive and Prescriptive Norms

Researchers could design experiments using descriptive norms that depart from prescriptive norms. For instance, they could place study participants in experimental scenarios in which they are told that the norm in their organization involves a discriminatory hiring practice (e.g., scanning online profiles to collect applicant demographics and then restricting job interviews based on certain demographics). At the same time, the study participant could be given a firm statement conveying a prescriptive norm regarding nondiscrimination and equal job opportunities. The goal would be to understand which norm, descriptive or prescriptive, affects the study participants’ top candidates for a position and the reasoning supporting their choices. These responses could be compared to conditions in which the study participant receives only the descriptive or the prescriptive norm, not both. This would provide a starting point for understanding the types of competing norm influence that individuals receive and allow researchers to examine the processes that differentiate the influences of norms that transmit behavior, perceptions, and prescriptions.

Beyond designing experiments that capture how norms compete, researchers could attempt to understand preexisting descriptive norms and how their influence shifts based on the transmission of prescriptive norms. One strength of Eriksson and colleagues’ studies was that the behaviors were fictitious, so the study participants could not possess past knowledge of swooshing or quining, which allowed the researchers to isolate the norm’s influence. This, however, does not represent real-world business practices. Employees may already engage in a specific behavior (e.g., gift giving) or hold a perception (e.g., gifts are not a form of unfair influence) that conflicts with a prescriptive norm (e.g., people ought not to give gifts when doing so constitutes bribery). This is an area in which researchers could provide impactful research. Organizations that want to promote ethical cultures need to understand the most effective way to communicate the prescriptive norm so that it can offset preexisting descriptive norms, especially those that are ingrained in an organization.

Beyond Norms Research

Several aspects of our analysis could be applied more broadly to behavioral ethics research that entails interventions that do not introduce information relevant for ethical reasoning. For instance, research has suggested that performing mathematical calculations causes unethical behavior (Wang, Zhong, & Murnighan, 2014; Zhong, 2011); that signing a form before, rather than after, completion reduces lying (Shu, Mazar, Gino, Ariely, & Bazerman, 2012); and that various forms of physical cleansing can remove ill effects associated with unethical conduct (Liljenquist, Zhong, & Galinsky, 2010; Zhong & Liljenquist, 2006). The authors assert that noncognitive processes underlie these effects. This stream of research has been critiqued from a variety of vantage points, including the inability to replicate findings in the same or different contexts (Berenbaum, 2021; Fayard, Bassi, Bernstein, & Roberts, 2009). Studies that rely on spurious shifts in ethical behavior without engaging in ethical reasoning are particularly important to recognize (Arkan, Nagpal, Scharding, & Warren, 2022). We believe that many of the issues that we note in terms of descriptive norms research could apply to a broader stream of behavioral ethics research in that meaningful shifts in ethical reasoning may not occur in these studies and may therefore limit our understanding of how to promote ethical organizations.

Role of Normative Scholars

Although we have addressed our discussion primarily to empirical researchers, our analysis offers guidance to normative scholars as well. In addition to making the case for why empirical scholars should include prescriptive concerns in their studies, our argument suggests a role for normative scholars to play in empirical research. Regarding the effectiveness-of-truth-telling-norm study suggested previously, for example, normative scholars are needed to establish with which ethical standards various norms conflict in particular scenarios. In studies that address descriptive norm conflicts with prescriptive norms, in turn, normative scholars are needed both to identify what is the prescriptive norm and to provide an account of what justifies its prescription.

While philosophers are not expected to be influenced by “what is” when determining “what ought to be,” we believe that normative scholars need to share responsibility for translating the ought to the is, which means that they need to engage in the descriptive research traditions by helping to guide empiricists in their studies. Just as we do not expect behavioral ethicists to dive into the philosophical literature to conduct an empirical study, we do not expect normative scholars to engage in empirical projects to ensure that business ethics studies include prescriptive norms. Yet, we believe our research shines a light on a much-needed role for philosophers as the behavioral ethics literature grows and takes hold with the research community and in real organizations.

In doing so, our research contributes to a small but important literature in business ethics that seeks not only to analyze differences between normative and empirical research (e.g., De Los Reyes et al., 2017; Singer, 1998; Treviño & Weaver, 1994; Weaver & Treviño, 1994) but also to incorporate normative scholarship into empirical research. Koehn (2020: 257), for example, has argued that recent trends in economics, such as behavioral economics, risk “creating harm or producing bad habits” unless they avail themselves of “ethical discernment.” Kim, Monge, and Strudler (2015), in turn, have argued that normative scholars should play a role in empirical research that appears to criticize normative theories, such as empirical accounts of “bounded ethicality.” This view holds that businesspeople are psychologically unable to fulfill many of the moral obligations that normative theorists seek to assign to them. Normative scholars are needed, according to Kim and colleagues, to explain what resources normative theories possess to respond to such empirical results and provide continuing normative guidance in light of them.

As an illustration of normative guidance for empiricists, we offer work by Alzola on virtue ethics. Alzola (2015) argues that the account of virtue favored by social scientists (what he calls the “reductive” account) is out of line with philosophical (“nonreductive”) accounts of virtue and subject to serious objections. Alzola’s argument suggests that it would behoove social scientists to include normative scholars in their empirical studies of virtue; in the absence of expert guidance in normative theories of virtue, empirical studies risk pursuing flawed research. Alzola’s argument also suggests, though, that normative scholars should seek to join empirical projects. By contributing to empirical projects, according to this logic, normative scholars can help to ensure that studies of ethics use the most rigorous conceptions of normative theories that have been developed.

To this growing literature in business ethics, our research adds specific tasks for normative scholars (e.g., establishing what are the prescriptions in empirical studies of prescriptive norms and what justifies them). We also affirm the views of Koehn (2020), Kim and colleagues (2015), Alzola (2015), and other scholars that normative theory is essential to empirical studies of ethics and, as such, that normative scholars should play a key role in empirical studies.

CONCLUSION

In this article, we emphasize the key difference between studies of different forms of norms and why it is important for research in behavioral ethics. Although many studies focus on how to encourage a community’s adoption of specific norms, they do not consider the prescriptive nature of the norm itself or, more importantly, what message is being transmitted. In many cases, the descriptive norm studies are merely replications of social proof theory, and the ethical significance of the norm (i.e., the study participants’ understanding of why the norm is ethical) is not addressed. Though it is useful to see that social proof occurs in many contexts, it is not clear that a prescriptive message is being transmitted or received. As such, the behavioral ethics research we have examined is vulnerable to the criticism that it merely extends social proof research in the domain of ethical behavior; it does not improve ethical reasoning and promote long-term ethical outcomes in organizations.