Neutralizing Unethical Negotiating Tactics: An Empirical Investigation of Approach Selection and Effectiveness

One of the fundamental truths of business is that effective negotiation is necessary for survival. Whether it is attracting customers, reducing the cost of raw materials, petitioning for a change in zoning laws, reworking a union management contract, or merging with a competitor, negotiation — the interpersonal decision‐making process undertaken whenever we cannot achieve our objectives single‐handedly (Thompson 2009) — is a central feature of business operations.

The process of engaging others to accomplish personal or professional objectives inevitably requires an exchange of information (expectations, positions, values, interests, options, etc.), which presents one of the major challenges facing negotiators (Butler 1999; Murnighan et al. 1999; Lewicki, Barry, and Saunders 2009). While sharing information can build trust and produce high‐quality outcomes, it can also put one at a disadvantage if the other party does not reciprocate (Roth and Murnighan 1982; Brodt 1994; Schweitzer, Hersey, and Bradlow 2006). At the very least, a negotiator would like to be able to discern when a counterpart is not being honest or forthcoming; under such circumstances, he or she could consider an alternative strategy, including aborting the negotiation.

Unfortunately, the use of questionable or unethical tactics in negotiations is far from an insignificant or rare occurrence (O'Connor and Carnevale 1997; Murnighan et al. 1999; Schweitzer and Croson 1999; Aquino and Becker 2005; Graebner 2009). Furthermore, humans often have difficulty detecting acts of deceit, even in face‐to‐face situations (Vrij 2000; Vrij and Mann 2004). And in this age of global markets, in which negotiations between diverse cultures often involve electronic exchanges, the process can be even more complex and challenging (Paese, Schreiber, and Taylor 2003).

A number of conditions are known to typically increase or decrease the likelihood that individuals will employ questionable or unethical tactics. Conditions that are likely to increase a counterpart's unethical tendencies — such as how important the negotiation is perceived to be (Tenbrunsel 1998; Robertson and Rymon 2001), how much time pressure the parties are under (Volkema and Fleury 2002), and the counterpart's reputation (Paese and Gilin 2000; Kickul 2001; Butt, Choi, and Jaeger 2005) — can be difficult to control in a negotiation.

The conditions that can decrease a counterpart's unethical tendencies, however, are often more manageable. For example, if a counterpart believes that long‐term business opportunities with the other party are possible, he or she will be more likely to enter into the negotiation with a harmonious disposition (Patton and Balakrishnan 2010) and employ soft (problem‐solving) rather than hard (competitive) strategies (Van Knippenberg and Steensma 2003). Likewise, explicitly noting linkages to a counterpart's professional network (e.g., contacts with a counterpart's business associates) can inhibit the use of questionable or unethical tactics, as the counterpart might fear that his or her behavior will be communicated to peers and others in the network.

According to Roger Volkema and Cheryl Rivers (2012), individuals communicate prospects for these and other conditions, consciously or subconsciously, to neutralize the tendencies of a counterpart to act unethically. In theory, these neutralizing behaviors can ultimately lead to a greater exchange of valid information, which should increase the likelihood and quality of an agreement. The relationship of such behaviors to the negotiation process, however, has been even less thoroughly examined than their impact on outcome.

In this study, we seek to examine the use and effect of twelve behaviors for neutralizing questionable or unethical negotiating tactics in a two‐party property leasing negotiation. As a first step, we examined the incidence and timing of these neutralizing behaviors. Next, we assessed the relationship of demographic and personality variables to the use of these behaviors. In addition, we examined the effect of these neutralizing behaviors on the negotiation process, focusing specifically on whether the initial use of neutralizing behaviors attenuated the counterpart's unethical tactics. Finally, we determined the relationships of these behaviors to actual and perceived joint outcomes and future negotiations.

Information is the currency of negotiation. When parties are able to freely exchange valid information, they are more likely to reach an integrative agreement (Thompson 1991; Olekalns, Smith, and Walsh 1996; Lewicki, Barry, and Saunders 2009). The primary risk of offering information about one's position, interests, resources, alternatives, deadline, and so on is that the other party might not reciprocate or might offer misleading or invalid information, resulting in a suboptimal outcome for one or both parties (Roth and Murnighan 1982; Brodt 1994).

A number of studies have reported on the extent to which ethically ambiguous negotiation tactics (EANTs) are employed in negotiations; these are tactics or maneuvers that might be regarded as improper depending on an individual's ethical reasoning and the circumstances, including everything from exaggerating demands and withholding information, to intentionally misstating facts and paying for inside information (Rivers and Lytle 2007). Kathleen O'Connor and Peter Carnevale (1997), for example, identified misrepresentations in 28 percent of negotiations. Maurice Schweitzer and Rachel Croson (1999) found that 25 percent of sellers both withheld information about their faulty products and lied about the conditions. Focusing on experienced negotiators in cases in which agreements were reached, Keith Murnighan and his colleagues (1999) found that 34 percent lied (i.e., made invalid statements) and were deceptive (i.e., failed to correct a counterpart's inaccurate assumptions). A higher incidence of unethical behavior was reported by Karl Aquino and Thomas Becker (2005) in a study in which individuals were offered an incentive to deceive: 43 percent of the participants concealed information, and 55 percent actively lied. Melissa Graebner's (2009) qualitative study of technology acquisition ventures revealed comparable numbers, with deception exhibited in more than 60 percent of the companies studied. Roger Volkema, Denise Fleck, and Agnes Hofmeister (2010) examined e‐mail negotiations in two‐party simulations, focusing on four tactics: exaggerating offers, misrepresenting information, pretending not to be in a hurry, and making promises that could not be kept. They found both parties used at least one of these tactics in two thirds of the negotiations, with information misrepresentation as the most common tactic. Collectively, these studies point to fairly widespread use of questionable or unethical tactics, which apparently can be spurred by situational factors (e.g., incentives, communication medium).

As suggested in several studies (cf. Pruitt 1981; Olekalns, Smith, and Walsh 1996; Adair and Brett 2005), individuals often begin with a competitive (distributive) orientation toward negotiating. That is, they often assume a “zero‐sum game” in which there is one “winner” and one “loser,” and employ self‐serving behaviors (e.g., exaggerated demands, affective persuasion) in an attempt to gain an advantage. This competitive approach can produce a similarly competitive, if not dishonest, response from a counterpart. For example, Wolfgang Steinel and Carsten de Dreu (2004) found individuals to be less honest and to engage in more deception when dealing with a competitive rather than a cooperative counterpart.

Given the importance of information to negotiations, the tendency of individuals to assume a competitive (win–lose) orientation at the outset of a negotiation, and the effects of such behavior on a counterpart's honesty, what can an individual do to keep a negotiation from spiraling into an exchange of misrepresentations and deception?

According to prospect theory (Tversky and Kahneman 1981), humans often are more motivated by potential losses than possible gains. Therefore, if a counterpart thought that his or her use of questionable or unethical tactics could endanger future business with this individual (Patton and Balakrishnan 2010) or negatively affect his or her network of associates (Brass, Butterfield, and Skaggs 1998), the counterpart might be reluctant to use such tactics.

A negotiating counterpart's potential costs or risks have been viewed along two dimensions. The first dimension is based on the dual‐concerns model for understanding conflict management behavior (cf. Thomas 1976; Rahim 1983; Ogilvie and Kidder 2008). The two concerns in this model are a concern for substantive outcome (i.e., fulfilling personal goals or accomplishing tasks) and a concern for relational outcome (i.e., the outcome for the other party and/or for the relationship). Thus, a counterpart may be reluctant to employ questionable or unethical tactics because of his or her concerns for goal/task accomplishment (Robertson and Rymon 2001; Schweitzer, Ordonez, and Douma 2004; Malshe, Al‐Khatib, and Sailors 2010) or concerns for interpersonal relations (Ben‐Yoav and Pruitt 1984; Curhan et al. 2008).

A second dimension that defines a counterpart's perceptions of risk is temporal orientation — short term versus long term. According to Yaacov Trope and Nira Liberman (2003), individuals' construal of objects and events depends on their temporal distance. They typically view near‐term events more concretely than distant events, which are seen as more abstract and central to the meaning of activities. Therefore, a counterpart might view risk differently, depending on when its consequences are most likely to be experienced.

According to Volkema and Rivers (2012), these two dimensions — goal/task concerns versus relational concerns, and short‐term orientation versus long‐term orientation — form four categories of risk that a negotiation counterpart might consider before using questionable or unethical tactics. These include risks to:

1. immediate or short‐term goal/task achievement (i.e., concerns about failure to reach agreement and unmet goals);

2. immediate or short‐term relationship(s) (i.e., loss of respect, support);

3. future or long‐term goals/task achievement (i.e., loss of future business, potential legal entanglements); and

4. future or long‐term relationships (i.e., impairment of future social/business networks).

These four categories, and sample neutralizing techniques for each category, are shown in Figure One and discussed below.

Because negotiators are typically motivated by their immediate or short‐term goals/tasks (Thompson 2009), providing assurances of goal accomplishment is one way to neutralize an unethical tactic (Robertson and Rymon 2001; Schweitzer, Ordonez, and Douma 2004; Malshe, Al‐Khatib, and Sailors 2010). Citing examples of prior successful negotiations (Fudge and Schlacter 1999; Kern and Chugh 2009), pointing out progress being made toward goal accomplishment (Rackham 2003), and noting the interdependence of the negotiating parties' goals (Rackham 2003; Comeau and Griffith 2005) also can focus a counterpart on pro‐social behavior rather than on questionable or unethical actions.

Interdependence is, to some extent, a function of the perceived availability of resources in a negotiation. A negotiating counterpart will likely consider the alternatives available to both parties when contemplating whether or not to use questionable or unethical tactics. According to resource dependency theory, counterparts who are aware of their lack of alternatives are more likely to avoid using unethical tactics for fear of being left without options should their tactics cause the other party to abort the negotiation (Kim, Pinkley, and Fragale 2005; Olekalns and Smith 2009). Likewise, a counterpart's belief that the other negotiator has alternatives also can impede his or her use of EANTs, as unethical behavior could prompt the negotiator to lose faith in the counterpart's trustworthiness and pursue a viable alternative. This was illustrated in a study by Marc Buelens and Dirk Van Poucke (2004), who found that when counterparts possessed information about negotiators' lack of alternatives, the counterparts significantly exaggerated their first offers to their benefit.

A perceived risk to short‐term relational interests also can deter a counterpart from employing questionable or unethical tactics. According to social identity theory, individuals develop a sense of self‐worth through membership in various social groups, often favoring others in those groups (Cialdini 1993; Yuki et al. 2005; Hargie et al. 2008). Studies also have found that individuals will employ fewer aggressive, self‐serving behaviors (Ben‐Yoav and Pruitt 1984) as well as less deception (Schweitzer and Croson 1999) with friends than with strangers. Thus, pointing out shared social identities (age, marital status, ethnicity, hobbies/interests, etc.) can decrease a counterpart's unethical tendencies.

Along this same line, getting a counterpart to identify with an ethical organization (e.g., pointing out the positive reputation of the counterpart's trade organization) can reduce the likelihood of him or her employing questionable or unethical tactics. This is comparable to Robert Cialdini's (1993) social proof principle, which involves referencing the desired behaviors of an individual's peer or aspirant social group. Studies of consumer behavior (Van Kenhove, DeWulf, and Steenhaut 2003) and whistle‐blowing (Taylor and Curtis 2010), for example, have shown that affective commitment will increase an individual's pro‐social behavior.

In addition, the counterpart's awareness of a negotiator's connections to his or her social network can affect the counterpart's decision to employ unethical tactics because the counterpart's use of such behaviors could more easily be communicated to significant others by the negotiator. According to social network theory, the greater the number and relational importance of individuals in one's social network who could learn about questionable or unethical behavior, the greater the probability that a negotiator will think carefully about reputational impact before acting (Brass, Butterfield, and Skaggs 1998). Roger Volkema and Maria Tereza Fleury's (2002) study of conditional ethics supports this theory, as they found participants less inclined to exaggerate an opening demand, misrepresent facts, and encourage others to defect to their side if they thought that their colleagues were more likely to learn the details of a negotiation.

Concerns for long‐term goals or tasks also can affect intentions and behavior, and vice versa. Game theory experiments, for example, have shown that individual behavior will become more competitive and duplicitous when players believe the negotiation is about to end (the end‐game phenomenon) (Axelrod 1984; Lount et al. 2008). In fact, Grainne Fitzsimons and Ayelet Fishbach (2010) found that negotiators can actually lose closeness to their counterparts as they approach goal completion. In contrast, when future interaction (and business) seem more likely, negotiators are often friendlier and more collaborative in order to protect future goals or opportunities (Volkema and Fleury 2002; Patton and Balakrishnan 2010).

If a counterpart perceives that his or her behavior may have legal implications, this also can serve as a deterrent (Bommer et al. 1987). Legal issues can take any number of forms, including fines, suits, license revocations, etc. (Gurley, Wood, and Nijhawan 2007), costs that may go well beyond the immediate losses of an unsuccessful negotiation (Menkel‐Meadow, Love, and Schneider 2006).

Finally, a counterpart may consider the risk of damaging long‐term relationships. Just as an individual might view a counterpart in terms of the assistance that he or she could provide in achieving future professional goals, an individual might seek to maintain a friendly, collegial relationship with his or her counterpart for purposes of future social support (Schneer and Chanin 1987; Peterson and Thompson 1997).

A counterpart's belief that a negotiator could provide access to networks of people who might offer relational support in the future may also exert a neutralizing effect on unethical tactics. A negotiator can stimulate this process by making reference to friends, associates, and social networks that he or she knows firsthand, suggesting affiliations that could be lost if a counterpart were to pursue a competitive approach involving questionable or unethical tactics.

Given these various techniques for neutralizing potentially unethical behavior, two research questions naturally arise:

1. Who is most likely to use these techniques?

2. What effect do the techniques have on negotiation process and outcome?

With respect to the first of these questions, the negotiation orientation or intentions of the individual employing neutralizing behaviors — integrative versus distributive — will help determine the expected relationship between personality variables and their use. For example, an individual could employ a neutralizing behavior (e.g., suggesting future business opportunities) as a means of keeping a counterpart from using questionable or unethical tactics (e.g., misrepresenting a price point), with the intention of seeking a mutually beneficial (integrative) outcome. Alternatively, an individual with a more distributive orientation, either initial or emergent, could recognize the potential for gaining an advantage in a negotiation by suggesting future business opportunities that he or she has no intention of offering or pursuing.

For purposes of hypothesis testing, we propose that over the course of a negotiation, negotiators will choose to use neutralizing behaviors that keep the negotiation on track in pursuit of a mutually beneficial outcome, which will also serve their self‐interests. Thus, we would expect that personality characteristics that have been positively linked in the literature to ethical decision making would likewise be positively linked to the use of neutralizing behaviors, while personality variables that have been negatively linked to ethical decision making would be negatively linked to the use of neutralizing behaviors.

In reviews of research on ethical decision making, several personality variables have been found to be associated with the use of such negotiation behaviors and tactics, including Machiavellianism, assertiveness, locus of control, risk propensity, self‐efficacy, and cooperativeness (cf. O'Fallon and Butterfield 2005; Craft 2013). Machiavellianism — an individual's tendency to pursue personal gain through self‐serving means — and assertiveness have been found to be negatively associated with ethical decision making (O'Fallon and Butterfield 2005; Barbuto and Moss 2006), the former both frequently and consistently. Some researchers have also found that individuals with an external locus of control, who believe that their destiny is controlled by outside forces, not themselves, are less likely to engage in ethical decision making than those who believe they control their own destiny (internal locus of control) (O'Fallon and Butterfield 2005); presumably one's perceived lack of situational control creates a justification for employing questionable or unethical tactics.

John Cherry and John Fraedrich (2002) found risk propensity to be negatively associated with ethical decision making, suggesting that risk‐taking individuals would be less likely to pursue ethical means (and more comfortable employing questionable or unethical tactics) than would risk‐averse individuals. Given the likely correspondence of neutralizing behaviors to ethical decision making, argued above, we might expect each of these four personality variables that have been found to be negatively associated with ethical decision making — Machiavellianism, assertiveness, locus of control, and risk propensity — to likewise be negatively associated with the use of neutralizing behaviors.

In contrast, cooperativeness has been found to be positively associated with ethical decision making (Robinson, Lewicki, and Donahue 2000). And regarding self‐efficacy, Rafik Elias (2009) found that the more an individual believed he or she was capable of accomplishing a goal or task, the more likely the individual was to behave ethically. As argued above, we might expect individuals high in self‐efficacy and cooperativeness to be similarly and positively inclined toward the use of neutralizing behaviors.

Therefore, our first study hypothesis is as follows:

Neutralizing behaviors can be used in a negotiation in two ways: as preventive measures (i.e., prior to either party's use of a questionable or unethical tactic) or as an intervention (i.e., in response to a counterpart's use of questionable or unethical tactics). Because the early stages of the negotiation process can play an important role in how the negotiation unfolds and the nature of the final outcome (Curhan and Pentland 2007; Volkema, Fleck, and Hofmeister 2011), preventive use of neutralizing behaviors can be particularly potent and crucial. Such behaviors provide assurances, for example, that progress is being made and that goals will be met.

In addition, these behaviors can nurture a positive tone or mood that is likely to affect similar, pro‐social behavior. Michael Hine and his colleagues (2009), for example, found that successful negotiations featured significantly more positive than negative words; negative language creates anxiety, which can increase the likelihood of employing questionable or unethical tactics.

Because a counterpart might be inclined to employ EANTs at various points in a negotiation, it may not be sufficient for a negotiator to use neutralizing behaviors only at the outset of an encounter — neutralizing behaviors might be necessary and effective at critical turning points in a negotiation to restore a collaborative focus. For example, reminding a counterpart of long‐term business opportunities or referencing one's social and professional contacts could potentially be more effective in moving the negotiation forward than countering with EANTs; neutralizing behaviors remind a counterpart of what could be lost if the negotiation turns contentious (Brass, Butterfield, and Skaggs 1998; Patton and Balakrishnan 2010). Thus, the use of neutralizing behaviors generally (i.e., proactively and reactively) could diminish a counterpart's overall use of unethical tactics.

Therefore, our second, third, and fourth hypotheses are:

Finally, we address the question of the effect of neutralizing behaviors on the outcome of a negotiation. Research suggests that competitive and unethical behavior in a negotiation is frequently reciprocated, which oftentimes leads to impasse (Axelrod 1984; Brett, Shapiro, and Lytle 1998; Boles, Croson, and Murnighan 2000). Because neutralizing behaviors are designed to reduce EANTs, which in turn should lead to the exchange of more valid information, we might expect that the greater the number of neutralizing behaviors, the more likely the parties will be to build trust and eventually reach an agreement. In addition, the improved exchange of valid information should improve the parties' joint outcome, both actual and perceived, because the negotiators will be more likely to discover and trade off issues that are important to one party but not to the other party, and vice versa (Tomlinson, Dineen, and Lewicki 2009).

Neutralizing behaviors may also have benefits in terms of future negotiations. Jochen Reb (2010) found that both the difficulty of the negotiating process and the perceived favorability of the outcome affected an individual's preference to negotiate with a counterpart again. As noted previously, many of the neutralizing behaviors (e.g., noting or referencing interpersonal similarities, goal progress, long‐term business opportunities, or personal/social support) are likely to encourage a positive climate or mood, at least compared with more contentious competitive and unethical tactics (Hine et al. 2009), positively influencing both the process and perceptions of the outcome. Thus, we might expect the use of neutralizing behaviors to positively affect an individual's desire to negotiate with a counterpart in the future.

Therefore, our fifth and final hypothesis is:

The participants in this study were two hundred thirty professionals from a large Brazilian financial institution. All participants were taking an online course on negotiation as part of a formal degree program. Drawn from states around the country, the participants had a mean age of 37.29 (standard deviation [SD] = 7.18), and 56 percent were male. At the time of the study, the course had not specifically addressed the topics of ethically ambiguous tactics or neutralizing behaviors.

Early in the course, participants were asked to complete a series of questionnaires. The questions included demographic information (e.g., age, gender) as well as five measures of personality — Machiavellianism, locus of control, risk propensity, self‐efficacy, and conflict style (which allowed for determination of assertiveness and cooperativeness). In each case, these questionnaires were translated into Portuguese and back to English to ensure an appropriate translation.

The Machiavellianism measure was based on the Mach IV questionnaire (Christie and Geis 1970), which has been used in prior negotiation research (cf. Amanatullah, Morris, and Curhan 2008). This questionnaire consists of twenty statements (e.g., “Never tell anyone the real reason you did something unless it is useful to do so”), each question rated on a 1 (strongly agree) to 7 (strongly disagree) scale. The Cronbach reliability estimate for the scale (α = 0.68) was adequate.

Locus of control questions were based on a questionnaire developed by Julian Rotter (1966). It consisted of ten statements (e.g., “My life is determined by my own actions”), four of which were reverse‐coded. Each statement was rated on a 5‐point scale, from strongly agree to strongly disagree. For this measure, the Cronbach alpha was 0.69.

The measure of risk propensity was determined by seven items adapted from Ree Meertens and Rene Lion's (2008) work, and three items adapted from Jinsook Cho and Jinkook Lee's (2006) research. For example, “Generally I view risks as a challenge” and “I am willing to take substantial risks to realize substantial financial gains from investments” represent sample items from the respective questionnaires. The ten statements were rated on a 9‐point scale (strongly disagree to strongly agree), with four of the statements reverse‐coded. The reliability estimate for this scale was 0.75.

Self‐efficacy was measured using ten statements from Ralf Schwarzer and Matthias Jerusalem's (1995) work; for example, “I can solve most problems if I invest the necessary effort” and “I can usually handle whatever comes my way.” Each statement is rated on a 1–4 scale (from not true at all to exactly true). The statements were adapted to have a negotiation focus, and had a Cronbach alpha of 0.78.

Measures for assertiveness and cooperativeness were derived from the Thomas–Kilmann Conflict Mode Instrument (Thomas and Kilmann 1974), which has been used in a number of studies involving conflict styles (cf. Ma 2007; Thomas and Thomas 2008). Assertiveness is commonly associated with self‐interested concerns (e.g., a competing style), while cooperativeness is associated with relational concerns (e.g., an accommodating style). According to Michael Chanin and Joy Schneer (1984), assertiveness and cooperativeness indices can be calculated from conflict styles scores for competing, collaborating, avoiding, and accommodating as follows:

These formulas were the basis for the assertiveness and cooperativeness measures in the study.

Subsequently, participants were assigned to one hundred fifteen dyads for an e‐mail‐based property leasing negotiation, with pairings randomly assigned by course administrators within regions but from different offices. In this negotiation, one party represented a telecommunications company (TELIB) interested in leasing office space, while the second party represented a property management company (IMOB). The negotiation involved eight issues — cost per square meter, renovation of rooms, utilities, length of lease, parking spaces, furnishings, advanced payment, and occupancy floor — with each issue assigned a set of outcomes with specific point values. With both integrative and distributive outcome potential, the negotiation is similar in format to simulations employed by other researchers (e.g., Olekalns and Smith 2007; Reb 2010). E‐mail messages were recorded through the course platform. The participants had ten days to complete the negotiation.

Two independent coders examined e‐mail transcripts for evidence of the twelve neutralizing behaviors shown in Figure One, as well as five questionable or unethical tactics from the Self‐reported Inappropriate Negotiation Strategies questionnaire (Lewicki and Robinson 1998; Ma 2010) that were deemed identifiable: making an exaggerated offer or demand, making promises that can't or won't be fulfilled, pretending not to be in a hurry, threatening to make your counterpart look weak or foolish in front of significant others, and intentionally misrepresenting factual information. (It was deemed that certain tactics, such as hiding one's real bottom line, could not be detected, while other tactics, such as talking to people to whom the counterpart reports or is accountable to encourage them to defect to one's side, could not occur in this online negotiation.) The inter‐coder reliabilities (Cohen's kappa) for the twelve neutralizing behaviors and the five tactics were 0.84 and 0.81, respectively. According to Roger Bakeman and John Gottman (1997), a kappa of 0.70 is considered to be in the “good” range. Any differences in coding were resolved through discussions.

Each participant's score for the negotiation was calculated as the sum of the points earned for each of the eight issues, with the integrative or joint score for the dyad calculated as the sum of the participant's score and his or her counterpart's score. In addition, each participant completed a postnegotiation questionnaire to determine perceived outcomes. Participants were asked how well they thought they had done in the negotiation (7‐point scale, where 1 = not well and 7 = very well), how well they thought their counterpart had done (7‐point scale, where 1 = not well and 7 = very well), and whether they would like to negotiate with their counterpart in the future (7‐point scale, where 1 = no and 7 = yes). The sum of the first two perceptions constituted perceived joint outcome.

To test Hypothesis One, we regressed the first use and overall use of the twelve neutralizing behaviors on Machiavellianism, locus of control, risk propensity, self‐efficacy, assertiveness, and cooperativeness (controlling for age and gender). For Hypotheses Two and Three, we also used regression analyses, controlling for the number of questionable or unethical tactics (EANTs) used in the same message as neutralizing behaviors, as some messages contained both neutralizing behaviors and EANTs.

We tested Hypothesis Four using correlation analysis. We examined the relationship between neutralizing behaviors and whether or not an agreement was reached (Hypothesis Five [a]) using logistic regression. Finally, we used regression analyses to determine the relationships between use of neutralizing behaviors and both actual and perceived joint outcomes (Hypotheses Five [b] and Five [c]) and the desire to negotiate with a counterpart again (Hypothesis Five [d]).

The one hundred fifteen dyads exchanged as few as two e‐mail messages and as many as twenty‐four messages (mean = 9.0 messages; SD = 3.7), with seventy‐one pairs (61.7 percent) reaching agreement. Overall, participants used three hundred fifty‐three neutralizing behaviors, with nearly two‐thirds (66.5 percent) employing at least one of the twelve behaviors. Participants used a mean of 1.53 (SD = 1.78) neutralizing behaviors. As shown in Table One, the most common neutralizing behaviors were suggesting that the negotiator had other alternatives (102 instances), suggesting that the parties' goals were linked (87), promising long‐term business opportunities (67), convincing the other party of goal progress (41), and suggesting limited options for the counterpart (27). The majority of neutralizing behaviors concerned goal interests rather than relational interests, with short‐term goal interests being the most common category of neutralizing behaviors.

A neutralizing behavior was used in the first e‐mail message by either party in sixteen of the negotiations. Of the behaviors used with some frequency, those behaviors used earliest in the sequence of e‐mail messages were suggesting limited options for the counterpart, getting the counterpart to identify with a respected organization, and suggesting that the negotiator had other alternatives. Convincing the other party of goal progress occurred latest in the exchange of e‐mail messages, around the eighth message on average. This timing is not surprising because a negotiation would logically require an exchange of issues and some tentative agreement before a comment on progress could be offered.

The means, SDs, and correlations for the primary independent, control, and dependent variables are shown in Table Two. Although we found no significant correlations for age, the results for gender suggest that, compared with males, females had a lower risk propensity (r = −0.18, p < 0.01), employed fewer EANTs per message (r = −0.15, p < 0.01), and had marginally lower joint outcomes (r = −0.15, p < 0.10). (Some of the dyads were single sex, some were mixed.)

The significant correlation between EANTs per message and neutralizing behaviors per message (r = 0.25, p < 0.001) suggests that neutralizing behaviors might be stimulating the use of EANTs, or vice versa. Dyads who communicated more frequently had better outcomes: the number of messages was positively associated with reaching an agreement (r = 0.35, p < 0.001), perceived joint outcome (r = 0.27, p < 0.001), and desire to negotiate in the future (r = 0.17, p < 0.05), while the number of EANTs per message was negatively associated with reaching an agreement (r = −0.17, p < 0.01), perceived joint outcome (r = −0.19, p < 0.01), and desire to negotiate in the future (r = −0.16, p < 0.05).

The results of the hierarchical regression analyses for Hypothesis One, shown in Table Three, reveal only marginally significant findings. In terms of first use of neutralizing behaviors, assertiveness was negatively related (beta = −0.12, p < 0.10), consistent with Hypothesis One (b). For overall use of neutralizing behaviors, only one personality measure was significant — risk propensity (beta = 0.13, p < 0.10) — and this was contrary to what we predicted in Hypothesis One (d), which was that a propensity toward risk would make someone less likely to use a neutralizing behavior. When broken down by the categories shown in Figure One, this finding for risk propensity is supported for behaviors in the short‐term goal/task category. For behaviors in the short‐term relational category, however, only external locus of control was significant: it was positively related to propensity to use neutralizing behaviors (beta = 0.12, p < 0.10), which was contrary to Hypothesis One (c).

The regression analysis for Hypothesis Two, which concerns the effect of the initial use of neutralizing behaviors on the use of EANTs, revealed statistical significance only for the control variable — EANTs used in conjunction with neutralizing behaviors. Specifically, the more EANTs a party used when he or she first employed neutralizing behaviors, the more EANTs a counterpart employed in replying to that e‐mail message (beta = 0.48, p < 0.001).

We also found no significant main effect for the use of neutralizing behaviors as a response to a counterpart's use of EANTs (Hypothesis Three). That is, the use of neutralizing behaviors did not significantly reduce the number of EANTs employed by a counterpart. In this case, the number of EANTs used in conjunction with neutralizing behaviors had no significant effect.

Overall, we found no significant correlation between one party's use of neutralizing behaviors and the other party's use of EANTs (Hypothesis Four). However, the use of neutralizing behaviors between the two parties was correlated (r = 0.34, p < 0.001), as was the use of EANTs between the two parties (r = 0.31, p < 0.001). In addition, we found a significant correlation between the use of neutralizing behaviors and the use of EANTs (r = 0.35, p < 0.001) for TELIB representatives, the prospective tenants, but not for IMOB representatives.

Given the results for Hypotheses Two through Four, we ran ad hoc correlation analyses to see if some neutralizing behaviors were more likely than others to be used in conjunction with EANTs. Focusing on the five most frequently used neutralizing behaviors, the following were significantly correlated with the simultaneous use of EANTs: suggesting that the negotiator has other alternatives (r = 0.24, p < 0.001), promising long‐term business opportunities (r = 0.17, p < 0.001), and suggesting limited options for the counterpart (r = 0.10, p < 0.001). Convincing a counterpart of goal progress and leading a counterpart to believe goals are linked were not significantly associated with the simultaneous use of EANTs. Further analysis revealed that neither of these two behaviors was used in an initial use of neutralizing behaviors (Hypothesis Two) nor did they significantly reduce the number of EANTs employed by a counterpart (Hypothesis Three).

With respect to Hypothesis Five (a), that greater use of neutralizing behaviors would increase the chances of the parties reaching an agreement, our logistic regression analysis found that the best predictor of agreement was the number of messages (beta = 0.29, p < 0.001). The more messages the parties exchanged, the higher the probability that they would reach agreement. We found no significant effects relative to Hypotheses Five (b) and Five (c), which concern the impact of neutralizing behaviors on actual and perceived joint outcome. But we did find that the number of EANTs per message (beta = −0.48, p < 0.01) and the number of neutralizing behaviors per message (beta = 0.18, p < 0.10) were significantly related to the parties' desire to negotiate in the future. That is, the more EANTs per message employed by the parties, the less they desired to negotiate again; the more neutralizing behaviors per message employed, the more they desired to negotiate again.

Negotiation is a process central to all forms of business. One of the challenges to negotiating effectively, particularly in a global marketplace involving diverse communication media, is how to deal with a counterpart who employs EANTs. Indeed, a number of studies have shown that anywhere from a quarter to two thirds of individuals will use questionable or unethical tactics during the course of a negotiation, depending on the situation or context. One way to neutralize this temptation is by employing behaviors that raise risks (i.e., potential costs to near‐ or long‐term goals or relationships) if questionable or unethical tactics are used. The purpose of this study was to investigate the use of twelve of these neutralizing behaviors in a simulated negotiation.

Of the one hundred fifteen dyads that participated in the e‐mail‐based property leasing negotiation, nearly two thirds of the participants employed at least one of the twelve neutralizing behaviors. The most common behavior was suggesting that the negotiator has other alternatives, with the most common category of behaviors being those that concerned short‐term goal interests. Neutralizing behaviors were used in a first message sixteen times, with most of these also falling into the short‐term goal/task category. The first use of a neutralizing behavior in a negotiation (i.e., before either party has sent a message containing one or more EANTs) is potentially revealing in a negotiation because neither party has yet to take a competitive stance. In this study, an individual's assertiveness, as measured through the Thomas–Kilmann Conflict Mode Instrument, was negatively related to first use, suggesting that the less assertive an individual, the more likely he or she was to use a neutralizing behavior in an initial e‐mail message.

First use of neutralizing behaviors, however, did not have a significant effect on the counterpart's subsequent use of EANTs (Hypothesis Two). Instead, the more EANTs used in conjunction with neutralizing behaviors, the greater a counterpart's use of EANTs (or, alternatively, the fewer EANTs used in conjunction with neutralizing behaviors, the fewer EANTs a counterpart employed). By combining neutralizing behaviors with EANTs in an e‐mail message containing the initial use of neutralizing behaviors, an individual may be sending a mixed message. In this case, it appears that the use of EANTs had the greater effect.

One of the inherent differences between many of the neutralizing behaviors and EANTs is that the former represent possibilities (e.g., future business opportunities, potential social contacts), while the latter (exaggerations, misrepresentations) represent real and immediate actions. As Roger Volkema, Denise Fleck, and Agnes Hofmeister (2011) suggest in a prior study of early‐stage behaviors, more credence may be given to actions (e.g., factual misrepresentations) than to words (e.g., the prospect of a mutually beneficial outcome or long‐term business opportunities). Therefore, practitioners seeking mutually beneficial outcomes must be careful in their messaging if they are indeed intent on getting the full benefits of employing neutralizing behaviors.

It seems logical that a number of the neutralizing behaviors — behaviors like getting a counterpart to perceive interpersonal similarities, promising a counterpart goal achievement, and promising a counterpart long‐term business opportunities — would be employed early in negotiations. These three behaviors, however, were used on average in about the sixth or seventh e‐mail message, while the average length of a negotiation was about nine messages. The deferred use of these neutralizing behaviors raises further questions regarding the intentions of their users: are these behaviors being employed for neutralizing purposes or as EANTs (e.g., feigning personal similarities or interest in long‐term business opportunities)? Further research is called for because the timing of behaviors has potential implications for both parties in a dyadic negotiation — the well‐intentioned initiator of neutralizing behaviors who may not recognize the implications of deferring use, and the counterpart who is using the timing of neutralizing behaviors as an indicator of intent.

The use of neutralizing behaviors as a response to a counterpart's use of EANTs also had no significant effect on a counterpart's subsequent actions (Hypothesis Three). In this case, the simultaneous use of EANTs also did not have an effect. It is conceivable that when a counterpart first employs EANTs and receives a mixed response of neutralizing behaviors and EANTs, he or she interprets the neutralizing behaviors as EANTs or, as noted above, gives more credence to the EANTs. The significant correlation between neutralizing behaviors and EANTs for TELIB representatives (Hypothesis Four) would support the theory that neutralizing behaviors were sometimes intended to act as EANTs. Some neutralizing behaviors are more likely than other behaviors to be used or interpreted in this way. For example, suggesting that one has other alternatives (the most frequently used neutralizing behavior in these negotiations; see Table One) could also be used or interpreted as a threat, and therefore as an EANT. Ad hoc analysis revealed that its use was significantly correlated with the simultaneous use of EANTs.

The number of neutralizing behaviors was not significantly correlated with EANTs for IMOB representatives, but the number of neutralizing behaviors was correlated for the two parties (Hypothesis Four). This could have been because the IMOB representatives viewed neutralizing behaviors differently from EANTs. The results for Hypothesis Five (d), in which the number of neutralizing behaviors was positively related to a desire to negotiate again with the other party, also point to neutralizing behaviors being viewed differently from EANTs. Future research might focus more specifically on goal‐related neutralizing behaviors, such as convincing a counterpart of goal progress and goals linkage, which were less likely to be used simultaneously with EANTs in this study. Not only are such behaviors nonthreatening, but a statement of goal progress speaks to the purpose of negotiating (goal achievement: Thompson 2009) while being easily verifiable by a counterpart.

These observations and findings, of course, must be interpreted within the context of the methodology of the study. First, we employed a single negotiation in this study. And while this negotiation had both integrative and distributive potential, it might be worthwhile to test these hypotheses under different conditions, varying the context (roles, purpose, time‐frame) and number of issues. Second, the participants in these negotiations had no guarantee or expectations that they would be negotiating together in the future. This perception may have limited the effects of certain neutralizing behaviors related to long‐term interests (e.g., suggesting links to a counterpart's social/professional network, proposing to be a gateway to valued social/business networks). It might also have focused participants more on the immediate task (which involved eight issues with point values) rather than relational issues. Third, this was an e‐mail‐based negotiation, and electronic mail is a medium that some scholars have suggested lends itself to deceptive and misinterpreted behavior (cf. Carlson et al. 2004). Therefore, it is possible that the participants in this study were more unethical than they would be in a face‐to‐face negotiation, and that the parties were quicker to interpret a counterpart's behavior as unethical.

Looking forward, future studies might employ other media and contexts to better understand these neutralizing behaviors. As noted, considerable research has shown that individuals often follow a relativist or contingent approach with respect to ethics (cf. Volkema and Fleury 2002). That is, a negotiator will often adapt his or her style and ethics to the actions of a counterpart. Particularly for negotiators with those personality types that are most likely to view questionable or unethical tactics as appropriate, we might presume a similar perspective regarding neutralizing behaviors that could be turned to one's advantage. The timing of use for specific behaviors might provide a further clue to whether or not neutralization for mutual or self‐benefit is the intention.

Several of the twelve neutralizing behaviors in this study were used infrequently or not at all. For the most part, these behaviors fell into the categories of short‐term and long‐term relational interests. The lack of use could be due to the communication medium employed in this study (electronic mail), a medium noted for its efficiency and with which there is often a tendency to move quickly into substantive discussions. This has potential implications for both practitioners and researchers. In terms of practitioners, making personal connections with a counterpart and his or her network of friends and associates can change the dynamics of a negotiation; individuals often negotiate more favorably with friends than with strangers (Greenhalgh and Chapman 1998). Negotiators may need to make a special effort to begin with relational conversation in order to take full advantage of these neutralizing behaviors. If a change in communication medium alone does not broaden the set of behaviors employed, it might be worthwhile for researchers to induce these behaviors in future experimental designs to more fully evaluate their effectiveness.