Gender Discrimination 1 GENDER DISCRIMINATION IN NEGOTIATOR DECEPTION: AN ARCHIVAL ANALYSIS OF BEHAVIOR IN THE MBA CLASSROOM

Gender Discrimination in Negotiator Deception: An Archival Analysis of Behavior in the MBA Classroom In this article, the relationship between gender and the likelihood of being deceived is examined. In strategic interactions, rational actors deceive on the basis of expected consequences (Gneezy, 2005). Gender stereotypes suggesting women are more easily misled than men predict that women will be deceived more so than men. Study 1 involved an archival analysis of deception in a real estate negotiation simulation in the MBA classroom (N = 298). Consistent with a rational analysis of gender stereotypes, women were deceived more frequently than men, yet women perceived no less honesty in their counterparts than did men. Economic and psychological consequences of deception were also examined, including agreement rates, sale price, and negotiator subjective experience. When believed by their target, lies facilitated deal making. However, psychologically, lying impaired subjective negotiation experiences. Studies 2 (N = 131), 3A (N = 107) and 3B (N = 132) were designed to explore the potential role of gender stereotypes in driving gender discrimination in negotiator deception. As expected, negotiator gender activated gender stereotypes, which in turn influenced the expected consequences of deception. Both theoretical and applied implications are discussed. Gender Discrimination 3 GENDER DISCRIMINATION IN NEGOTIATOR DECEPTION: AN ARCHIVAL ANALYSIS OF BEHAVIOR IN THE MBA CLASSROOM “... Salesmen ... categorize people into "typical" buyer categories. During my time as a salesman I termed the most common of these the "typically uninformed buyer".... [In addition to their lack of information, these] buyers tended to display other common weaknesses. As a rule they were indecisive, wary, impulsive and, as a result, were easily misled. Now take a guess as to which gender of the species placed at the top of this "typically easy to mislead" category? You guessed it-women.” (Parrish, 1985 p. 3, as quoted by Ayres & Siegelman, 1995) The reformed car salesman’s quote above reveals a truth about social perception: expectations about the vulnerabilities and strengths of interaction partners are shaped in part by their gender. Shared, category-based expectations, or stereotypes, exist about one gender versus another (Fiske, 1998). Though the activation of stereotypes is automatic and unavoidable (Devine, 1989), how stereotypes are applied to guide behavior in particular situations is often at the discretion of social actors. In competitive negotiations like car purchases, one party’s gain is another’s loss. As such, self-interested negotiators seeking to secure attractive deals may rely on gender stereotypes to make strategic decisions. The current research examines the relationship between gender stereotypes, the expected ease of being misled, and the decision to deceive men versus women in negotiations. Gender Discrimination 4 Gender stereotypes create a broad set of challenges for women negotiators. Because effective negotiators are thought to possess stereotypically masculine traits such as assertiveness and rationality (Kray & Thompson, 2005), negotiating often means acting counterstereotypically for women. Burdened with unflattering stereotypes, women negotiators profess more anxiety, less knowledge about the process, and less confidence in their ability (Babcock, Gelfand, Small, & Stayn, 2006; Kray & Gelfand, 2009) compared to their male counterparts. As a result, women are vulnerable to the debilitating performance effects of negative stereotypes (Kray, Thompson, & Galinsky, 2001), typically performing worse than men at the bargaining table (Stuhlmacher & Walters, 1999). In addition to producing differences in how men and women approach negotiations, gender stereotypes also affect the manner in which women and men are treated. Stereotypes create a set of expectations about how individuals should behave and those who fail to live up to them often experience social repercussions (Rudman, 1998). Indeed, prescriptive gender stereotypes dictating that women be agreeable may lead female negotiators to be judged more harshly for the identical behaviors of male negotiators (Bowles, Babcock, & Lai, 2007). Using a vignette, Bowles and colleagues manipulated both whether a job candidate attempted to negotiate a job offer and the candidate’s gender. Attempting to negotiate reduced participants’ willingness to work with female candidates relative to male candidates who engaged in identical negotiation attempts. Gender Discrimination 5 The effect of gender stereotypes on negotiator behavior has also been shown in naturalistic settings, including the car dealership. In a striking field study, Ayres and Siegelman (1995) had women and men actors follow a standardized script inquiring about a new car purchase at various auto dealerships. They found that price quotes were significantly higher for women than for men. Because this discrimination occurred irrespective of salesperson gender, the researchers argued it was not simply a result of prejudice against women (which would presumably be more evident by male salespersons than female salespersons). Instead, they argued, discrimination resulted from a “statistical inference” being drawn by salespeople whereby gender-based cues determined the expected profitability of deals and resulted in disparate price quotes between the sexes (Phelps, 1972). Because the trained actors adopted identical bargaining strategies, it is likely that gender stereotypes, rather than behavioral differences, led salespeople to expect women were more likely than men to pay a high markup. In the words of the researchers (p. 317), “If sellers believe, for example, that women are on average more averse to bargaining than men, it may be profitable to quote higher prices to women customers.” The current research extends this work on gender discrimination, or behavior aimed at denying particular social groups positive outcomes (Allport, 1954), and bargaining to the domain of deception. By examining whether feminine stereotypes imply that women are easier to mislead and, if so, whether women negotiators are especially likely to be misled, theory and research are extended on multiple fronts. First, the relationship between deception and counterpart gender in negotiations is examined for the first time. In his Gender Discrimination 6 groundbreaking research on deception in strategic interactions, Gneezy (2005) did not study gender differences and subsequent work has only examined gender from the perspective of the liar (Dreber & Johannesson, 2008). Second, the current work extends our understanding of gender stereotypes relevant to behavior at the bargaining table to examine assumptions about negotiators’ ease of being misled. Third, by utilizing a fully crossed design with respect to dyad gender composition, this research overcomes methodological limitations of previous work in this domain. In so doing, a more comprehensive picture of the role of gender in negotiations emerges. Finally, by exploring negotiation behavior in the MBA classroom, the applied implications of gender in negotiations are explored. DECEPTION IN NEGOTIATION Deception, or communication aimed at intentionally misleading another person, is often driven by self-interest (DePaulo et al., 1996). Because self-interest is a guiding force in negotiations, it is not surprising that deception is prevalent (Lewicki, 1983; Schweitzer & Croson, 1999). Though men are more willing to engage in unethical negotiating tactics than women (Dreber & Johannesson, 2008; Lewicki & Robinson, 1998; Robinson, Lewicki, & Donahue, 2000), whether gender influences the likelihood of being deceived in negotiations remains largely unexamined. Beyond the negotiation domain, DePaulo et al.’s (1996) study of lying in everyday life sheds light on the role of interaction partner sex. Using a diary methodology recording spontaneous lies, interactions involving women were found to involve more “white lies” Gender Discrimination 7 (i.e. meant to protect their feelings). Yet because a wide range of interaction and relationship types were included in this analysis, this finding may not fully reflect women’s avoidance of competition (Niederle & Vesterlund, 2007). In situations with an inherently competitive element (i.e. negotiations), it remains an open question whether lies aimed at providing a strategic advantage are told more frequently to one gender versus the other. By holding constant the context involving a decision of whether to lie, the current research is able to examine whether highly competitive, strategic interactions invite gender bias. GENDER STEREOTYPES AND THE DECISION TO DECEIVE Why would negotiators conclude that the risks associated with lying are lower with female negotiators? To address this question, consider the ethical decision making process. In strategic interactions, expectations about consequences guide the decision to deceive (Gneezy, 2005). In other words, rational actors consider both the subjective probability of getting caught and the cost of punishment. Viewed through the lens of gender stereotypes, both considerations put women at a disadvantage relative to men. With respect to lie discovery, feminine stereotypes involving agreeableness and even gullibility suggest women are less likely than men to recognize and to challenge a lie. With respect to punishment, masculine stereotypes involving aggressiveness suggest men are more likely than women to retaliate against a discovered lie. If women are deemed less “risky” targets of deception than men, as the introductory quote suggests, then they become vulnerable to being deceived. Gender Discrimination 8 Gender stereotypes suggest, first and foremost, that women must be warm and nice (Bem, 1974; Prentice & Carranza, 2002). This feminine imperative to be agreeable conflicts with the simple act of negotiating, leading women negotiators to appear pushy and demanding (Amanatullah & Morris, 2010; Bowles et al., 2007; Small, Gelfand, Babcock, & Gettman, 2007). In general, attempting to negotiate can be considered impolite. Because conforming to politeness norms is particularly important for low status individuals such as women (Brown & Levinson, 1987), women demonstrate reluctance to initiate negotiations (Small et al., 2007) and dislike the negotiation process more so than men (Kray & Gelfand, 2009). Prescriptive feminine stereotypes demanding niceness may also render women reluctant to accuse another of lying or, minimally, to lead their interaction partners to expect this reluctance. Accusatory reluctance is characterized by a discomfort in labeling others as deceptive (Ekman, Friesen, O’Sullivan, & Scherer, 1980; O’Sullivan, Ekman, Friesen, & Scherer, 1985). Almost by definition, the warmth and kindness expected of women mandates hesitancy in accusing others of foul play. Doing so is unpleasant, uncomfortable, and potentially aggressive, all of which violate prescriptive feminine stereotypes. In addition to the mandate that women be agreeable, the fact that women are allowed to possess certain undesirable traits, such as being gullible, impressionable, and naïve (Prentice & Carranza, 2002), may increase women’s likelihood of being deceived. These relaxed proscriptions for women provide a means of avoiding the uncomfortable task of Gender Discrimination 9 accusing another of lying. As in the case of whistleblowers, accusing others of wrongdoing often carries penalties that people are motivated to avoid (Johnson, 2002). Furthermore, to the degree that women are presumed to possess these undesirable characteristics, a rational actor might conclude women are also more likely to believe a lie. Indeed, women’s own admission of their lack of knowledge about negotiating (Kray & Gelfand, 2009) likely lowers the subjective probability of their catching a lie at the bargaining table. Overall, the subjective risks of getting caught in a lie appear to be lower with female negotiators. Another potential consequence of deception is the threat of retaliation should the lie be discovered. Examined from this angle, women remain disadvantaged relative to men. Masculinity is associated with agency and aggression (Williams & Best, 1982); these masculine stereotypes suggest men are more likely than women to retaliate against a discovered lie. Once again, aside from any actual behavioral differences between men and women in retaliation, awareness of gender stereotypes may affect expectations about retaliation, thus influencing a rational actor’s decision to deceive. Likewise, even if women were as likely as men to retaliate, their ability to do so effectively may be limited. Women’s ability to punish a liar may be relatively constrained because women have lower status than men (Jackman, 1994) and status predicts the amount of attention granted by others (Fiske, 1993; Keltner, Gruenfeld & Anderson, 2003). Each of these considerations leads to the prediction that women are more likely than men to be deceived in strategic interactions. Gender Discrimination 10 OVERVIEW OF STUDIES Four studies were conducted to determine the relationship between gender, gender stereotypes, and deception in negotiations. Study 1 was designed to explore deceptive behavior in the context of a negotiation simulation in the MBA classroom. Though the simulation was hypothetical, behavior in the simulation could have real consequences both in terms of reputations and for ongoing relationships with classmates. The central question is whether women are deceived more often than men and, if so, what are the actual consequences of doing so? The remaining studies were conducted to explore a possible underlying mechanism driving gender discrimination in deception, the role of gender stereotypes in shaping expectations about negotiators. To this end, Study 2 examined the link between gender and expectations. It was hypothesized that female negotiators would be expected to be more easily misled and less tough than male negotiators. Studies 3A and 3B examined the relationship between gender stereotypes and the expected consequences of negotiator deception. It was hypothesized that feminine stereotypes would be associated with positive consequences of deception (i.e. ease of reaching agreement) whereas masculine stereotypes would be associated with negative consequences of deception (i.e. threat of retaliation). STUDY 1: DECEPTION IN THE MBA CLASSROOM If gender stereotypes lead negotiators to expect women to be more easily misled than men and these expectations shape ethical decision making, then female negotiators should be deceived more often than their male counterparts. To test this hypothesis, an archival dataset was created using existing measures from an MBA negotiation course 1 . Gender Discrimination 11 In this course, students completed face-to-face negotiation role playing exercises followed by post-negotiation online surveys on a weekly basis. Though the economic terms of deals were not graded, preparation and effort in the exercises were. In addition to this performance incentive, students were also highly motivated to reach attractive deals. After each exercise, the precise terms of each negotiating pair’s agreement were summarized in written form and shared with the entire class, thus providing clear and immediate reputational incentives to do well. As a result of this debriefing process, students could be virtually assured that any deception occurring during negotiations would ultimately be revealed to their negotiating partners. Finally, negotiators could be assured of ongoing interactions with their counterparts, with the possibility of negative reputational and relational consequences of deception extending beyond the simulation (cf. Anderson & Shirako, 2008). One exercise in particular was designed to introduce the concept of ethics, thus providing an ideal context for testing the hypothesized relationship between gender and deception. The negotiation task (described in detail below) involved a buyer-seller real estate transaction designed to pose the following ethical dilemma to buyers: do they lie about their intended use of the property to facilitate a deal that might not otherwise occur? 2 Prior to debriefing the exercise (i.e. revealing the buyers’ true intentions), participants completed a post-negotiation survey including all dependent variables. One measure of ethical decision making was buyer self-reported lying, a common method for assessing various forms of dishonesty (cf. DePaulo et al., 1996; McCabe, Gender Discrimination 12 Butterfield, & Trevino, 2006). Because this self-report measure may have been influenced by the perceived consequences of lying, it was supplemented by open-ended descriptions of the buyer’s tale (i.e. their stated intentions), provided by both buyers and sellers. Two independent judges coded these descriptions for deception. A subset of the sample (due to pedagogical reasons) also assessed the buyer’s degree of honesty, which provided both a finer-grained sense of buyer behavior and a measure of whether sellers detected buyer deception. Finally, consequences of deception were examined to assess whether the expectations derived from gender stereotypes predicted negotiation processes and outcomes. Economic consequences included agreement rate and sale price. Psychological consequences included dyad ratings of their negotiating experience. In contrast to the hypothesized positive economic consequences of deception for buyers, it was expected that lies (even if undetected) would impair negotiators’ psychological experiences (Sagarin, Rhoads, & Cialdini, 1998). Because women sellers were expected to be deceived more than male sellers, two consequences were hypothesized: 1) more positive economic terms were expected for buyers negotiating with female sellers; 2) more negative negotiating experiences were expected for dyads including female sellers. METHOD Participants. Participants were 298 full-time M.B.A. students at a public west coast business school (221 male, or 74.2%) who were enrolled in one of six sections of a negotiation course, comprising 149 dyads (65 male-male, 23 female buy-male sell, 48 Gender Discrimination 13 male buy-female sell, 13 female-female) 3 . Both male and female students were randomly assigned to negotiation roles. Given that men comprised approximately 75% of MBA enrollment, data from 6 sections of negotiation classes across 3 semesters were combined to enable the analysis of a full factorial design 4 . The negotiation exercise occurred in approximately the 4 th week of a 15-week course. Procedure. Participants were given one hour to negotiate the “Bullard Houses” role playing exercise (DRRC version, 1995). Participants were randomly assigned to negotiate as the buyer’s agent (“buyer”) or the seller’s agent (“seller”) in a real estate negotiation. As in past research examining ethical decision making (Kern & Chugh, 2008), this simulation was chosen because it provides negotiators with a range of options for responding to an ethical dilemma posed to buyers. Sellers were instructed only to sell the property to a known, reputable buyer for “tasteful” and preferably residential purposes. Buyers were prohibited from revealing under any circumstances that their client intended to build a commercial high-rise hotel on the property catering to tourists and convention visitors, a use inconsistent with the sellers’ interests. However, at no point were buyers instructed to lie. Buyers were faced with the decision of whether to be truthful versus dishonest to sellers about their client’s intended use of the property. Buyers exhibiting total honesty could inform the seller that they were prohibited from revealing the intended use of the property, though doing so may raise suspicion and thereby potentially increase the risk of reaching an impasse. Alternatively, various degrees of dishonesty could be employed: Gender Discrimination 14 buyers could claim that they were unaware of their client’s intended use, focus on ambiguous terms like “residential” (though a hotel “houses” people, it is short-term and requires different zoning than long-term residences), or blatantly lie by claiming that their clients intended to put the property to a use consistent with the seller’s interests (i.e. brownstones). Dependent Measures Lie Admissions. Lie admissions were measured by asking buyers if they lied to sellers about the intended use of the property, coding no as “0” and yes as “1.” Buyer Deception. Buyers’ deceptiveness was independently assessed in two ways: 1) buyers’ open-ended responses to: “What did you tell the seller about the intended use of the property?” and 2) sellers’ open-ended responses to: “What is your understanding of the intended use of the property by the buyer?” Both buyer and seller descriptions were coded for deception by two independent judges on the following 6-point scale: 0 (truth, i.e. “high-rise commercial” or “buyer said he was not authorized to reveal”), 1 (truth but violated client’s orders, i.e. “hotel”), 2 (vague and subjective, emphasis on “tasteful” use), 3 (claimed ignorance or uncertainty about purpose, i.e. “undecided”), 4 (misleading, emphasis on “residential” use), 5 (blatant lie, i.e. “condominiums”). Coder reliability was adequate for both buyer ( = .69) and seller ( = .65) descriptions, so judges’ ratings