Predicting Competitive‐Unethical Negotiating Behavior and Its Consequences

The negotiation process is central to managerial and organizational success (Mintzberg 1973; Thompson 2005). A key to mastering this process is the ability to manage the exchange of information between the parties (Thompson 1991; Shell 1999). When negotiators are able to freely share information regarding their issues, positions, and interests, an agreement satisfactory to all parties is more likely to be reached (Thompson 1991; Olekalns, Smith, and Walsh 1996; Lewicki, Barry, and Saunders 2007).

Negotiators face the dilemma of how much information to share and how quickly. If the parties know each other well and have negotiated successfully in the past, they may share a level of trust that encourages one or both to share confidential or insider information (e.g., issue importance, reservation price, alternatives) (Butler 1999). If the parties lack this trust, however, they may withhold information for fear that the other party will not reciprocate, thereby putting the initiating party at a disadvantage (Roth and Murnighan 1982; Brodt 1994; Schweitzer, Hersey, and Bradlow 2006).

To nurture familiarity and trust, negotiators often attempt to identify commonalities between the parties (e.g., interests, experiences, colleagues) and build rapport through expressions of approval or support, proximity, and compliments (Cialdini 1993; Thompson, Peterson, and Kray 1995; Cialdini 2009). Studies have shown that individuals are generally more comfortable sharing information with friends and like‐minded persons than with strangers (cf. Valley, Neale, and Mannix 1995; Halpern 1996).

Negotiators may also, however, employ tactics designed to gain an advantage in a negotiation (Boles, Croson, and Murnighan 2000; Olekalns and Smith 2007). False promises and misrepresented information, for example, may create a sense of rapport and trust that will lead the target party to reveal confidential information. This deception can be difficult to detect, requiring an ability to discern and integrate a number of behavioral cues and contextual factors (Vrij 2000; Vrij and Mann 2004).

Several researchers have reported incidents of unethical behavior in negotiations. Kathleen O'Connor and Peter Carnevale (1997), for example, found that participants in a laboratory study misrepresented information in 28 percent of their negotiations. Keith Murnighan and his colleagues (1999) found that 34 percent of experienced negotiators both lied (i.e., made invalid statements) and were deceptive (i.e., failed to correct a counterpart's inaccurate assumptions) in cases in which negotiations led to an agreement. Terry Boles, Rachel Croson, and Keith Murnighan (2000) also reported individuals to be deceptive about 13 percent of the time in their laboratory study, with more deception occurring in the early stages of negotiation.

Several potential predictors of competitive and unethical behavior have been suggested, including:

• economic and performance pressures on the negotiator (Husted 1999; Volkema and Fleury 2002);

• demographic (e.g., age, gender) and cultural characteristics (Husted 1999; Volkema 2004), including the lack of salient ethical standards (Aquino 1998) or a sanctioning system (Tenbrunsel and Messick 1999);

• a counterpart's reputation for benevolence (Olekalns and Smith 2007), temptation (Tenbrunsel 1998), or dishonesty (Volkema and Fleury, 2002); and

• the medium employed, that is, media with low to moderate information richness, such as electronic mail, producing lower levels of pre‐negotiation trust than face‐to‐face encounters (Naquin and Paulson 2003).

On the other hand, direct questions (Schweitzer and Croson 1999) and negative‐ethicality feedback (Kim, Diekmann, and Tenbrunsel 2003) have been found to curtail unethical behavior once the negotiation has begun.

While researchers have sought to measure and compare the attitudes of individuals toward various competitive and unethical behaviors (Lewicki and Robinson 1998; Robinson, Lewicki, and Donahue 2000; Volkema 2004) and to relate various personality traits to moral reasoning (Rayburn and Rayburn 1996; Moberg 1999; Forte 2005), much less is known about the relationship of attitudes toward competitive or unethical behavior and actual behavior, and the impact of that behavior on the process and outcome. The use of competitive‐unethical tactics could significantly influence both actual and perceived outcomes along with the reputations of the parties, particularly if such behavior occurs in the early stages of a negotiation when trust is not yet established (Volkema, Fleck, and Hofmeister‐Toth 2004; Schweitzer, Hersey, and Bradlow 2006).

This article reports on a study of negotiators' attitudes toward competitive and unethical tactics and the affect of those attitudes on initial and overall behavior of participants in a two‐party, e‐mail‐based negotiation. In addition, the relationships between competitive‐unethical behavior and perceptions of performance and honesty/reputation (which, in turn, can influence subsequent negotiations) are examined. Finally, the implications of these findings for negotiators and future research are discussed.

The issue of ethicality in negotiation requires an appreciation for the range of tactics or behaviors that negotiators might employ during an encounter. Following on the work of Sissela Bok (1978) and others, Roy Lewicki (1983) developed a typology of questionable or unethical tactics consisting of five categories:

1. traditional competitive bargaining (e.g., exaggerated first offers),

2. information misrepresentation (e.g., misrepresenting facts, progress),

3. bluffing (e.g., making promises one will not deliver),

4. questionable information collection (e.g., feigning friendship for information), and

5. influencing a counterpart's professional network (e.g., discrediting a counterpart with his/her superiors).

Importantly, this work recognized that tactics are not easily categorized as simply ethical or unethical but rather lie on a continuum of perceived appropriateness or acceptability. Traditional competitive bargaining tactics, for example, often are perceived to be more appropriate or acceptable than the tactics found in the other four categories (cf. Robinson, Lewicki, and Donahue 2000; Volkema 2004), though still capable of giving the individuals employing those tactics an advantage in a negotiation (Dees and Cramton 1991).¹

Because the use of competitive‐unethical tactics is likely to affect the tone if not the outcome of a negotiation (discussed in more detail later in this article), it is important to understand what factors might predict the use of these tactics. As previously noted, a number of variables have been suggested as potential predictors, including the personality of the negotiator (Rayburn and Rayburn 1996; Moberg 1999; Forte 2005), the circumstances or conditions of the negotiation (Volkema and Fleury 2002; Ross and Robertson 2003), and the perceived behavior of a negotiator's counterpart (Tenbrunsel 1998; Paese and Gilin 2000; Butt, Choi, and Jaeger 2005; Olekalns and Smith 2007).

In terms of personality, research on conflict resolution suggests that some individuals are more concerned with self‐interests than with the interests of their counterparts or the relationship (Thomas 1976; Rahim 1983). More recent research on bargaining and negotiation has found that individuals can be oriented more toward distributive (win–lose) outcomes than integrative (win–win) outcomes, or vice versa (Lax and Sebenius 1986; Bac 2001; Nelson and Wheeler 2004). As a consequence of these preferences, we might expect differential attitudes toward the negotiation tactics and behaviors that are employed to achieve those outcomes.

A wide variety of studies have examined the linkages between individual orientation/attitudes, intentions, and actual behavior, dating back to the early work of Icek Ajzen and Martin Fishbein (1969, 1970, 1972; Ajzen 1988). In a meta‐analysis of these and other studies, Min‐Sun Kim and John Hunter (1993) found statistical support for linkages between attitudes and intentions, attitudes and behavior, and intentions and behavior for a cross‐section of behavioral choices (e.g., political candidates, automobiles, cultural events). In cases in which an individual's actions fell within his or her control, intentionality was also found to be a significant mediating variable between attitudes and behavior (Kim and Hunter 1993; Chang 1998).

Although these studies did not examine ethical attitudes or intentions per se, it is reasonable to expect similar linkages for negotiations because most competitive‐unethical tactics fall within a negotiator's control (e.g., exaggerating an initial offer, misrepresenting information). Indeed, several studies have found strong correlations between attitudes toward these tactics and intentions to use them (cf. Lewicki and Robinson 1998; Volkema 1999b). The linkages between attitudes and behavior also seem likely, given the more generic findings (Kim and Hunter 1993). Thus, we have developed our first hypothesis.

We might expect to find not only a significant link between attitudes toward competitive‐unethical behavior and the quantity of corresponding tactics employed, but the haste with which that behavior occurs as well. Recall that Boles, Croson, and Murnighan(2000), in their study of the dynamics of deception and retribution in repeated ultimatum bargaining between dyads, found participants to be more deceptive in the early stages of a negotiation than in the later stages. Game theory research involving the prisoner's dilemma has demonstrated that the orientation of a participant — cooperative, individualistic, or competitive — frequently plays out in the initial round of the game (Deutsch 1973). Thus, it seems likely that the more favorably an individual views competitive or unethical tactics, the more likely he/she would be to use competitive‐unethical tactics early in the negotiation. The rationale for such behavior could be to gain an immediate advantage (e.g., by anchoring the negotiation with an exaggerated initial offer) (Thompson 2005) or to preempt a counterpart's anticipated competitive‐unethical behavior (Volkema and Fleury 2002). The latter — preemptive behavior — can be based on either internal attributions (i.e., the presumed unethical nature of a counterpart) or external attributions (e.g., a highly competitive negotiating environment), which leads to our second hypothesis.

Generally speaking, individuals respond in kind to another party's actions or deeds (Cialdini 1993). In negotiations and conflict situations, this reciprocation has been found to include matching integrative and distributive communications, procedural statements, and affective statements (Donohue 1981; Putnam 1983; Weingart et al. 1990; Brett, Shapiro, and Lytle 1998; Paese and Gilin 2000; Adair and Brett 2005; Butt, Choi, and Jaeger 2005). Even when the parties do not communicate, as in simulations such as the prisoner's dilemma, tit‐for‐tat behavior is not uncommon (Axelrod 1984).

Individuals often feel that they have a right, if not an obligation, to protect themselves against the potential risks and costs of maintaining ethical standards when their negotiating counterparts behave dishonestly or seem untrustworthy (Dees and Cramton 1991). Further, we expect an individual with an orientation toward competitive‐unethical behavior would require little prompting to employ one or more of these tactics: a counterpart's initial exaggerations or deceit would conjure thoughts of worst‐case scenarios (Parks, Sanna, and Posey 2003; Zhang and Han 2007) and provide justification for responding in kind (Brett, Shapiro, and Lytle 1998; Volkema and Fleury 2002). Thus, just as a perceived honest disclosure will produce less exaggeration (Paese, Schreiber, and Taylor 2003), an indication of competitive or unethical behavior would, we expect, be likely to lead to the corresponding behavior (e.g., exaggeration, misrepresentation), particularly if attributions shift to internal factors (the fundamental attribution error). We have developed our third and fourth hypotheses accordingly.

Taken collectively, these hypotheses suggest that individuals who are intent on gaining and maintaining an advantage in a negotiation are going to employ competitive or unethical tactics early and often. Suspected by their counterparts, it will likely be viewed as a signal of aggressive, self‐serving intentions (Sinaceur and Neale 2005; Olekalns and Smith 2009), producing reciprocation of a similar if not more unethical nature between the parties. Therefore, our fifth hypothesis is:

In general, the more frequently a negotiator exhibits a particular behavior or category of behaviors, the greater the likelihood that those behaviors will be recognized and he/she will begin to develop a reputation (Schweitzer, Hersey, and Bradlow 2006). Because individuals have different views regarding the appropriateness of negotiation tactics (Robinson, Lewicki, and Donahue 2000; Volkema 2004), both the quantity and quality of tactics employed are likely to affect pattern recognition and, ultimately, a counterpart's assessment (e.g., honest, dishonest). In addition, the choice of communication medium can mediate this process, as face‐to‐face encounters allow for detection of an individual's nonverbal cues while a medium like electronic mail produces a record of prior messages that enables a negotiator to check for inconsistencies and affords him or her some time between responding to messages to search online databases for contradictory or confirmatory information (Paulson and Naquin 2004; Vrij and Mann 2004; Schweitzer, Hersey, and Bradlow 2006).

While the use of competitive‐unethical tactics is often most effective when their intent or nature is undetected, behavioral repetition increases the probability that patterns will be recognized and suspicions raised (Carlson et al. 2004). Affected parties can share their observations (and conclusions) with colleagues who will be cued to look for similar patterns in contemporaneous and future encounters (Tinsley, O'Connor, and Sullivan 2002; Volkema and Fleury 2002). Again, a medium such as electronic mail with a record of prior correspondence allows for additional opportunities for detection, that is, by back‐checking statements, offers, and promises.

Given these factors, we offer our sixth hypothesis.

In addition to behavioral frequency, early behavior (primacy effect) has been found to more greatly influence impression formation in a variety of contexts than has later behavior (recency effect) (cf. Anderson 1965; Park 1986). Within the context of negotiation, this also appears to be the case. Jeffrey Rubin and Bert Brown (1975), for example, reported a negotiator's early behaviors to be more instrumental than his/her later behaviors in shaping the way the individual is perceived. Marwan Sinaceur and Margaret Neale (2005), examining implicit and explicit threats offered both early and late in the negotiation process, found that explicit threats offered early in a negotiation led to the highest levels of perceived aggressiveness. But even when an individual's questionable or unethical behavior does not occur in the initial exchange, a dramatic change in behavior can produce a noticeable effect (Cialdini 1993, 2009). Therefore, our seventh hypothesis involves the initiation of behavior.

Finally, while laboratory studies of bargaining and negotiation generally measure the relative outcomes of participants, real‐life negotiations seldom afford the parties an opportunity to gather comparative outcome data from their counterparts or an omniscient observer. Instead, each negotiator bases his or her satisfaction with an agreement, and his or her willingness to do business with the other party in the future, on a subjective assessment (Oliver, Balakrishnan, and Barry 1994). This relative perceived performance can be an important factor in both the implementation of the agreement and in the negotiation of future agreements (Curhan, Elfenbein, and Xu 2006).

A negotiator's perceptions of outcome can be based on a number of factors, not the least of which is the extent to which he or she believes that the agreement was formed using valid and complete information (Paulson and Naquin 2004). If the negotiator believes that he or she has been honest and also believes that his or her counterpart has been honest during the negotiation, then the negotiator likely will suppose that the agreement was formed using complete and valid information. Further, the more complete the information exchange, the greater the likelihood of an integrative outcome (Roth and Murnighan 1982; Butler 1999), which leads to our eighth hypothesis.

Further, given the proposition that actual competitive‐unethical behavior will be negatively related to perceived honesty (Hypothesis Six), we suggest our ninth hypothesis.

Sixty‐six graduate students from two business negotiation courses participated in the study. The focal negotiators in this study had a mean age of 25.6 years, and 60.6 percent were male. The language for the simulation (described below) was English, which was also the primary language used in the two courses. (While not all students were native English speakers, all were fluent.)

The participants were asked to assume roles in a two‐party, property‐leasing negotiation (Volkema 1999a; Volkema, Fleck, and Hofmeister‐Toth 2004). The thirty‐three students from one class represented a telecommunications company called Logan Telecommunications, which sought to lease commercial property to expand its international operations. Specifically, Logan Telecommunications was interested in one or more properties managed by RJW Properties, Inc., which was the role taken by the thirty‐three students from the second class. Negotiating dyads were randomly assigned and remained the same throughout the study. Because the students simulating Logan Telecommunications were from a different class than the students taking on the RJW Properties roles, we can assume that the students typically did not know their negotiation counterparts prior to the simulation. In addition, negotiation ethics had not been discussed or otherwise addressed in class prior to the simulation.

The parties were asked to negotiate seven issues: cost per square meter, duration of lease, advanced payment, renovations, furnishings, utilities, and parking space. Each issue had a predetermined set of outcomes, and each outcome produced a specific point value that differed for Logan Telecommunications and RJW Properties. The scoring tables were communicated via written background information. As is typical with multi‐issue tasks of this type (cf. Thompson 1991; Olekalns, Smith, and Walsh 1996), the point values allowed for integrative as well as distributive outcomes. More specifically, there were three issues with high and medium‐low point values, and also issues for which each party could achieve the same or nearly the same outcome as his/her counterpart but only at his/her counterpart's expense. The scenario also included two issues more important to Logan Telecommunications than to RJW Properties and two issues more important to RJW Properties than to Logan Telecommunications. Thus, distributive as well as integrative outcomes were possible. This simulation had previously been used successfully in face‐to‐face negotiation simulations.

All negotiations took place via electronic mail, which has emerged as an important medium for conducting domestic and international business negotiations (Shell 2001; Nowak 2003). Generally used in asynchronous communications, e‐mail is substantially more economical than face‐to‐face meetings, particularly for the growing number of international negotiations that organizations are now engaged in. It also has the advantage of easy storage and retrieval of text‐based messages (McGinn and Croson 2004). As a medium with only moderate information richness, however (Daniels 1967; Daft and Lengel 1986), electronic mail can require lengthier negotiations, particularly when integrative outcomes are possible. Consequently, participants were allotted two weeks to complete their negotiations (which also allowed sufficient time for rumination and fact checking).

At the conclusion of the negotiations, we collected transcripts from participants. Then, Logan Telecommunications representatives were asked to complete a questionnaire to determine how well they thought they had done and how well they thought the other party had done (seven‐point scale, 1 = not well to 7 = very well). They were also asked to rate how honest they had been and how honest the other party had been (seven‐point scale, 1 = not very honest/ethical to 7 = very honest/ethical). (No prenegotiation measures of a counterpart's perceived ethicality were taken, as this could have tipped off participants to the research agenda and affected their expectations and behavior during negotiations.)

To test Hypotheses One through Three, we needed to measure the attitudes of the participants toward various negotiation tactics. We did this using a questionnaire developed by Roy Lewicki and his colleagues (Lewicki 1983; Lewicki and Robinson 1998; Robinson, Lewicki, and Donahue 2000) that has been used successfully in a number of studies of negotiation and ethics (cf. Volkema and Fleury 2002; Perry and Nixon 2005). The questionnaire, called the “Incidents in Negotiation Questionnaire,” focuses on eighteen tactics that represent a range of behaviors from competitive bargaining behaviors to more questionable tactics such as misrepresenting information and manipulating another party's professional network (Anton 1990; Lewicki and Robinson 1998). Each behavior is rated on a seven‐point Likert scale in terms of its appropriateness (1 = not at all appropriate, 7 = very appropriate) and likelihood of use (1 = not at all likely, 7 = very likely). The questionnaire indicates that there are no “right answers.” Because of the high correlation found between these two measures in earlier studies (cf. Lewicki and Robinson 1998; Volkema and Fleury 2002) as well as this study (r = 0.70, p < 0.001), only the appropriateness measure was employed in this study. As with other studies of this type (Olekalns and Smith 2007), the questionnaire was administered several weeks in advance of the negotiation.

The competitive‐unethical tactics actually employed by participants were identified by three individuals (two professors and a graduate assistant) working independently and with no prior knowledge of the negotiated outcomes. An estimate–discuss–estimate approach was employed (Nutt 1992), with differences resolved through discussion. The three individuals were asked to identify four tactics: exaggerated offers, misrepresented information, pretending not to be in a hurry, and making promises that could not be kept. These were the only four tactics from the eighteen found in the “Incidents in Negotiation Questionnaire” that could realistically occur and be measured in the context of this particular simulation. (Other tactics either involved third parties, such as paying others for information, going to a counterpart's supervisor, or engaging the press, or were difficult to measure, such as hiding one's bottom line.) These four tactics represent three of the five categories of competitive‐unethical tactics identified by Lewicki and Robinson (1998): competitive bargaining, misrepresenting information, and bluffing. The number of competitive‐unethical behaviors (Hypotheses One, Four to Six, and Nine) was the sum total of uses of these four behaviors by a negotiator (Eagly and Chaiken 1993).

For purposes of this study, an exaggerated offer was defined as a value beyond the range specified in the scoring table. For example: “We are willing to pay $400/square meter” (when the specified range was $500–$900); “Our company needs places for ten cars” (specified range was zero to four cars); lessee requested “refrigerator/stove/ice machine” (although only refrigerator and stove were specified in the table). Misrepresented information was defined as a statement that went beyond the factual background information provided to participants (e.g., “We have grown at a rate of 50 percent for the last three years,”“I saw the rooms of your properties and they are in really bad condition,” and “Most of our lessees pay six months in advance”).

Pretending not to be in a hurry was a special category of misrepresenting information that involved giving a false sense of unavailability or casualness (e.g., “I really can't negotiate on Saturday . . . because the university is closed for the weekend”). Making promises that could not or would not be kept included promises that fell beyond the purview of the negotiator (e.g., “If we see that RJW is reliable, we certainly will make future business with you”). Taken from actual transcripts, these sample statements were likely to be viewed by recipients as dubious within the context of the negotiation (i.e., questionable given the ranges specified in the scoring tables, the common background information provided each party, the uncertainty of future business opportunities) or through Internet research (e.g., university academic calendars).

All hypotheses were initially tested using correlation analysis, with the exception of Hypothesis Two (which we tested using discriminant analysis). Follow‐up regression analysis was employed to identify the most significant predictors of competitive‐unethical behavior. Because of the nonindependence of negotiating dyads, all hypotheses except Hypothesis Four were analyzed from the point of view of a single member of the negotiating dyad — the Logan Telecommunications representative — referred to as the focal negotiator (Kenny, Kashy, and Cook 2006).²

At least one of the two parties used a competitive or unethical behavior in thirty‐one (93.9 percent) of the thirty‐three negotiations. In twenty‐two cases (66.7 percent), both parties used a competitive‐unethical behavior. The most commonly used tactic was misrepresenting information (twenty‐five times by Logan Telecommunications representatives, twenty‐two times by RJW Properties representatives), followed by exaggerating an offer or demand (ten times by Logan representatives, fifteen times by RJW Properties representatives). Both Logan Telecommunications and RJW Properties representatives employed competitive or unethical tactics in initial e‐mail messages ten times (30.3 percent of all negotiations).

Twenty‐two (66.7 percent) of the thirty‐three pairs of negotiators reached agreement. In twenty‐one of the twenty‐two cases in which an agreement was reached, at least one of the parties used a competitive or unethical behavior. In ten of the eleven cases in which no agreement was reached, at least one of the parties used a competitive‐unethical tactic. There was no significant correlation between the use of competitive‐unethical behaviors and reaching an agreement (r = 0.09, p = 0.62); however, there was a marginally positive correlation between reaching an agreement and the perceived honesty of the other party (r = 0.32, p < 0.10).

Overall, Logan Telecommunications negotiators' attitudes toward the eighteen competitive‐unethical behaviors (the sum of the eighteen ratings), as measured by the “Incidents in Negotiation Questionnaire,” ranged from 36.00 to 99.00 (mean perceived appropriateness = 66.82, standard deviation = 16.92) (alpha = 0.89). The most acceptable behavior of the four focal tactics was pretending not to be in a hurry (mean = 5.94, standard deviation = 1.39), followed by exaggerating offers (mean = 5.52, standard deviation = 1.44), misrepresenting information (mean = 3.85, standard deviation = 2.15), and making promises that could not be kept (mean =  2.15, standard deviation = 1.37). These values were comparable to those reported elsewhere (cf. Volkema 1997).

We found the focal negotiators' attitudes toward competitive‐unethical tactics to be predictive of their actual use of such tactics (Table One). We found a positive correlation between perceived appropriateness of competitive‐unethical behaviors and the number of such tactics the negotiator employed (Hypothesis One: r (correlation) = 0.49, p (probability) < 0.01). (Note: As shown in Table One, attitudes/perceived appropriateness was calculated for all eighteen tactics as well as just the four focal tactics. Because these sums were highly correlated, r = 0.77, p < 0.001, analyses proceeded with only the broader attitudinal measure.) In addition, discriminant analysis revealed that the perceived appropriateness of competitive‐unethical behavior was predictive of whether the negotiator used one of those tactics in an initial e‐mail message (Hypothesis Two: canonical correlation = 0.35; Wilkes' lambda = 0.88; p < 0.05).

We found no significant correlation between attitude/perceived appropriateness and the number of messages delivered between a counterpart's initial competitive‐unethical behavior and the focal individual's competitive‐unethical response (Hypothesis Three). A number of the other hypotheses regarding timing and frequency of use of competitive‐unethical behaviors, however, were statistically significant.

The more frequently an individual employed competitive‐unethical behavior in his/her e‐mail negotiation, the more frequently his or her counterpart employed competitive‐unethical tactics (Hypothesis Four: r = 0.42, p < 0.05). If, for example, a Logan Telecommunications representative did not use any competitive‐unethical behaviors, his or her RJW Properties counterpart employed about two such tactics on average. If a Logan Telecommunications representative used five competitive‐unethical behaviors, however, his/her counterpart also employed five such tactics on average.

As expected, the earlier in the negotiation that an individual employed competitive‐unethical behavior, the more competitive‐unethical behaviors in total he or she used during the negotiation (Hypothesis Five: r = −0.47, p < 0.01). Upon further examination, the data were found to best fit a logarithmic model: number of tactics employed = −2.06 Log (first use) + 4.79 (F = 10.43, p = 0.003). Thus, the earlier that competitive or unethical tactics were employed, the more it amplified the overall use of such behavior. This is consistent with Sinaceur and Neale's (2005) finding that early threats produce heightened aggressiveness.

Because several variables were found to correlate significantly with competitive‐unethical behavior, we conducted regression analysis to see which of these variables were the strongest predictors of such behavior. In this analysis, the independent variables were: attitude toward competitive‐unethical tactics (i.e., perceived appropriateness), the party initiating competitive‐unethical behavior, and a counterpart's use of competitive‐unethical behavior, as well as pair‐wise interaction effects. The dependent variable was the number of competitive‐unethical tactics employed. We found two independent variables to be significant predictors of use: an individual's attitude toward competitive‐unethical tactics and the interaction of an individual's attitude toward those behaviors and the competitive‐unethical behavior of his or her counterpart (Table Two). Specifically, the number of tactics employed = 0.060 (attitude toward tactics) + 0.004 (attitude toward tactics × number of tactics employed by counterpart) – 1.984 (F = 11.23, p < 0.001). The interaction term suggests that attitude and reciprocity can be a particularly potent combination, as the more appropriate a focal negotiator perceived the use of competitive‐unethical tactics and the more such tactics were employed by his or her counterpart, the more competitive‐unethical tactics the focal negotiator employed.

While the correlation was in the predicted direction, we found no significant relationship between the number of competitive‐unethical behaviors actually employed by RJW Properties representatives and Logan representatives' perceptions of the honesty of their RJW Properties counterparts (Hypothesis Six: r = −0.26, p = 0.15). But when we broke the number down by competitive tactics (i.e., exaggerated offers, pretending to be in no hurry) and unethical tactics (misrepresenting information, making false promises), we found the number of behaviors falling into the former category to be significant. Specifically, the number of competitive behaviors employed by RJW Properties representatives was negatively correlated with Logan Telecommunications representatives' perceptions of their RJW counterparts' honesty (r = −0.42, p < 0.05). Because exaggerated offers were defined in this study as values beyond the range shown in the scoring table, there was little doubt as to their validity or the RJW Properties representatives' intent — it was clear to counterparts that these offers were exaggerated. Further, this finding suggests that even exaggerated offers (at least in the extreme) were considered not very honest or ethical.³

Not surprisingly, the correlation between the number of competitive‐unethical behaviors used by Logan representatives and their self‐perception of honesty was statistically significant (r = −0.52, p < 0.01). Because Logan representatives were well aware of the intent of their own use of unethical behaviors (i.e., misrepresenting information, making false promises), the correlation for these was more significant (r = −0.48, p < 0.01) than was the correlation for competitive behaviors (r = −0.36, p < 0.05). Still, as noted above, competitive behaviors were perceived as not very honest or ethical.

We found a marginally significant statistical relationship between a party's use of a competitive‐unethical tactic first in the negotiation and the counterpart's view that the negotiator was less honest than the counterpart (Hypothesis Seven: r = 0.32, p < 0.10) (not shown in Table One). When broken down by competitive tactics and unethical tactics, the correlations were statistically significant for the former (r = 0.53, p < 0.05) but not the latter (r = 0.15, p = 0.47). The apparent certainty of exaggerated offers in this study likely left little doubt in the minds of Logan Telecommunications representatives as to RJW Properties representatives' early intentions, which appear not to have been the case for tactics like misrepresenting information and making false promises. When differential perceived honesty was regressed against both first use and differential use of competitive‐unethical tactics, however, it was the latter that proved to be the more significant predictor: differential perceived honesty = −0.20 (differential use of competitive‐unethical tactics) + 0.82 (F = 5.90, p < 0.05).

In terms of perceived outcomes, the greater the perceived collective honesty of the parties (according to Logan Telecommunications representatives), the greater the joint outcome was perceived to be (Hypothesis Eight: r = 0.52, p < 0.01). A follow‐up analysis revealed that the perceived honesty of the RJW counterpart was more positively correlated with perceived joint outcome (r = 0.55, p < 0.001) than was the Logan representative's perceived honesty (r = −0.03, p = 0.89). But we found no significant correlations between collective perceived honesty and joint actual performance, or between perceived joint performance and actual performance. Finally, we found no significant difference between the parties' use of competitive‐unethical behaviors and how well Logan representatives perceived they performed in these negotiations (Hypothesis Nine).

Competitive and unethical behaviors appear to occur fairly commonly in dyadic negotiations. In this study, at least one competitive‐unethical behavior was employed by each party in 66.7 percent of the negotiations, numbers considerably larger than the percentages reported in previous studies involving face‐to‐face negotiations (28 percent, O'Connor and Carnevale 1997) and electronic negotiations (13 percent, Boles, Croson, and Murnighan 2000; 38 percent, Murnighan et al. 1999). The higher incidence may be an artifact of the relatively moderate information richness of the medium employed (asynchronous e‐mail), which lends itself to lower levels of prenegotiation trust (Naquin and Paulson 2003) and ultimately to higher levels of deception (Paulson and Naquin 2004). It could also reflect the fact that the negotiating dyads in this study lacked prior contact or the potential for future encounters. Negotiators often will employ more aggressive, self‐serving behavior with individuals whom they do not know or do not expect to see again (Ben‐Yoav and Pruitt 1984). Although not a specific focus of this study, these moderating effects deserve further investigation.

Overall, our results support the hypothesized link between participants' attitudes toward competitive‐unethical negotiating tactics (i.e., perceived appropriateness) and their actual behavior. The attitudes of the focal negotiators in this study, as measured by the “Incidents in Negotiation Questionnaire,” proved to be good predictors of both the quantity (number) of competitive‐unethical tactics employed (Hypothesis One) and the use of a competitive‐unethical tactic in a first e‐mail message (Hypothesis Two). While this questionnaire has been used for comparative purposes in a number of studies (for a review, see Rivers and Lytle 2007), these results offer initial support for the predictive validity of the questionnaire. These findings are consistent with the larger and more general body of research first conducted by Ajzen and Fishbein (1969, 1970, 1972; Ajzen 1988) and subsequently reaffirmed by others (cf. Kim and Hunter 1993; Chang 1998) connecting attitudes with behaviors.

The only lack of significant findings regarding Logan Telecommunications participants' attitudes was with their responses to their RJW counterparts' initial competitive‐unethical behavior (Hypothesis Three). This result might be due in part to the coding system that was employed for the number of e‐mail messages between a counterpart's initial competitive‐unethical behavior and a competitive‐unethical response. In those cases in which there was no competitive or unethical response, the response was coded as one more than the responding individual's final message (rather than infinity or indeterminate). However, it is also possible that the timing of a negotiator's competitive‐unethical behavior might be better predicted by his/her counterpart's behavior than by the negotiator's ethical orientation (Volkema and Fleury 2002).

Regression analysis, which we used to help identify the best predictor(s) of competitive‐unethical behavior, revealed an important finding. The best predictor of competitive‐unethical behavior was the interaction of two variables: the attitude of a focal negotiator toward competitive‐unethical behavior and the behavior of his/her counterpart. The more appropriate a focal negotiator perceived competitive‐unethical tactics to be and the more his or her counterpart employed these tactics, the more such tactics the focal negotiator employed. In addition, the focal negotiator's attitude toward competitive‐unethical tactics was positively related to their use. The significant interaction effect suggests that situational factors (i.e., a counterpart's behavior) can fuel a negotiator's behavior, ethical or unethical, consistent with what has been predicted by others (cf. Trevino 1986; Volkema and Fleury 2002).

Taken in combination with other findings regarding reciprocity and early use of competitive‐unethical behavior, a favorable attitude toward competitive‐unethical tactics can prove to be a self‐fulfilling prophecy. Early use of competitive‐unethical tactics leads to greater frequency of use (Hypothesis Two, Hypothesis Five), which in turn increases the chances of reciprocal behavior (Hypothesis Four). Reciprocal behavior provides justification for one's initial intentions and behavior, which then leads to an escalation of competitive‐unethical behavior. The logarithmic relationship between early use and frequency of use is an interesting finding because it means that use of competitive‐unethical tactics in first and second messages can compound the behavior. Furthermore, being a first user of competitive‐unethical tactics in a negotiation can create a stigma that lasts throughout the negotiation (Hypothesis Seven), although differential use by the parties may be a better predictor of perceived relative dishonesty.

On the other hand, these findings also suggest that a counterpart's choice not to reciprocate (especially to the early use of tactics) can actually decrease the other party's use of competitive‐unethical tactics, although the other party's initial behavior may produce some residual use. This, of course, can be a difficult approach to take, particularly if the other party signals a distributive orientation by employing competitive‐unethical tactics in his or her very first message.

As might be expected, the focal negotiators in this study were more accurate in assessing their own honesty than the honesty of their counterparts. The difficulty in assessing a counterpart's deception is not uncommon (Vrij 2000; Vrij and Mann 2004), particularly when the medium has moderate information richness, such as electronic mail (Paulson and Naquin 2004; George, Marett, and Tilley 2008). But in this study exaggerated demands were defined as offers that went beyond the range specified in the scoring table (comparable to a negotiator knowing the true market value of a counterpart's offer) and misrepresented information went beyond the factual background information provided to participants, so these were not only readily detectable but more likely conceived of as self‐serving. Consequently, focal negotiators' perceptions of their counterparts' honesty was significantly correlated with the use of this competitive behavior.

Finally, this study examined the relationship of ethicality to actual and perceived outcomes. Until recently, the latter — perceived or subjective outcomes — had received little attention by researchers. Jared Curhan and his colleagues (2006) however have argued that subjective outcomes are important in negotiations: while laboratory studies often compare objective measures of performance, in reality negotiators seldom have access to their counterpart's utility functions to make such a comparison. Instead, it is subjective feelings regarding the process and outcome that determine a negotiator's satisfaction.

The larger the perceived collective honesty of the parties, the greater the joint outcome was perceived to be (and, conversely, the smaller the perceived collective honesty, the smaller the perceived joint outcome). The correlation of perceived collective honesty to joint actual performance, however, was not significant. Interestingly, it was the focal negotiator's perceived honesty of his or her counterpart in this study that seemed to be the better predictor of how well the individual thought the parties did collectively in the negotiation. Thus, focal negotiators appear to have discounted their own competitive‐unethical attitudes and behaviors, attributing consequences to perceptions of a counterparts' competitive‐unethical behavior (which was likely seen to trigger comparable behavior). This is consistent with the tendency of individuals to attribute success to internal factors (e.g., personal ability, motivation) and to attribute failure to external factors (Schilit 1990).

The implications of these findings, of course, must be understood within the context of the study's limitations. First, this was a negotiation simulation wherein the participants had no relational history and no apparent opportunity to engage their counterparts in future negotiations. In this regard, the simulation comes closer to a first encounter between individuals completing a one‐time deal. As a consequence, the parties would not have used competitive‐unethical tactics for defensive purposes in an initial e‐mail message based on their counterparts' unethical reputations — because there were no prior negotiations on which to base reputations. Future research examining the moderating effects of different conditions — prior business encounters (positive, negative, nonexistent) and future business opportunities (unlikely, possible, certain) — on the use of competitive and unethical behavior (Ben‐Yoav and Pruitt 1984) would be useful.

Second, we examined the use of competitive‐unethical behavior in terms of the frequency of using four tactics/behaviors that could logically occur in the case simulation and feasibly be measured: pretending not to be in a hurry, exaggerating offers, misrepresenting information, and making promises that could not be kept. While these would seem to be somewhat representative of the range of such tactics, it would be worthwhile to examine other tactics in future research. In addition, if a valid weighting of each behavior's ethicality could be determined, it might be worthwhile to go beyond both count data and breakout categories (competitive, unethical) to assess degree of ethicality. As previously noted, this is challenging, because appropriateness appears to vary according to demographic and situational factors. Third, some tactics are harder to sense or identify than other tactics, particularly when the individuals who employ competitive or unethical tactics are doing their best to obscure their intentions. We chose to count tactics that had a high probability of being recognized as questionable or dubious (e.g., an offer that went beyond the range of options, a claim that went beyond common background information). While negotiators will sometimes lay traps to determine their counterpart's veracity (e.g., securing inside information, then playing dumb), it might be worthwhile in future studies to examine competitive‐unethical tactics that are more difficult to discern (e.g., feigning disinterest).

Finally, electronic mail, a medium of moderate information richness, was employed in this study. While organizations have come to rely more and more on electronic mail for domestic and international negotiations (Shell 2001; Nowak 2003), the medium has a number of potential limitations compared with face‐to‐face negotiations, including lower levels of prenegotiation trust (Naquin and Paulson 2003). Researchers have reported that electronic negotiations typically feature less rapport, less information exchanged, more threats and ultimatums, and more impulsive, confrontational, and negative behavior (Thompson and Nadler 2002; Valley et al. 2002; Paese, Schreiber, and Taylor 2003), which can result in a greater likelihood of distributive outcomes and impasses than found in face‐to‐face negotiations (Valley, Moag, and Bazerman 1998; Loewenstein et al. 2005). Therefore, some caution should be exercised in extending the findings of this study to negotiations employing other media or combinations of media.

With the continued globalization of world markets, however, we might expect the relevance of these potential limitations to diminish. That is, more and more companies are undertaking negotiations with new business partners, although partners with whom they hope to have future business relations. In addition, global competition continues to force organizations to employ more convenient, economical communication media such as electronic mail for their negotiations. Consequently, understanding the relationships between ethical attitudes, behavior, and consequences and the use of these media will become increasingly important.