Acknowledging the Other Side in Negotiation

Theorists and practitioners alike have long recognized the value of fostering a cooperative context for negotiation (Johnson and Johnson 1994), one in which participants engage in joint problem‐solving exercises (see Kelman 1972; see also Doob, Foltz, and Stevens 1969; Cohen et al. 1977; Kelman and Cohen 1986; Neale and Bazerman 1991; Kelman 1993; Rouhana and Kelman 1994). In particular, experts urge negotiators to seek mutual gains through integrative bargaining (Fisher, Ury, and Patton 1991; see also Raiffa 1982; Susskind and Cruikshank 1987; Thompson 1998) that creates value or “enlarges the pie,” instead of simply struggling to win the largest possible slice of that pie (Walton and McKersie 1965; Lax and Sebenius 1986).

Beyond emphasizing the importance of cooperative norms in fostering integrative bargaining, some investigators have given specific attention to interpersonal and affective aspects of negotiation (see Carnevale and Pruitt 1992; Pruitt and Carnevale 1993; Valley, Neale, and Mannix 1995; Greenhalgh and Chapman 1996; Bazerman, Curhan, and Moore 2004; Beersma and De Dreu 2005; Curhan, Elfenbein, and Xu 2006). In one early example of this approach, Alan Benton, Harold Kelley, and Barry Liebling (1972) found that experimental participants faced with an opponent who was initially extreme in his demands but subsequently proved willing to compromise felt more responsible for (and more satisfied with) the negotiation outcome than participants who faced an opponent who made less extreme demands from the outset but thereafter remained intransigent.

The results obtained by Benton and his colleagues support our more specific contentions about the importance of attributional considerations and the role of acknowledgment in the give‐and‐take of the negotiation process. In particular, recipients of a compromise proposal typically must come up with answers to the dual questions of “Why this concession?” and “Why now?” before accepting the proposal. By framing an offer as expressly designed to accommodate the position of the other side, concession offerers may go a long way toward solving the attributional dilemma for the recipient (i.e., “This particular concession is being offered to me now because it is a response to my expressed concerns”; see Baron 1985; see also Ross and Ward 1995).

Researchers in many areas of psychology other than bargaining and negotiation have long recognized that enhancing an individual's feelings of efficacy or control (even if that control is illusory) can significantly improve that individual's sense of well‐being and everyday functioning (e.g., Glass and Singer 1972; Sherrod 1974; Bandura 1977; Langer 1983). And influential work by Tom Tyler (1987) and E. Allan Lind (e.g., Lind and Tyler 1988) has suggested that gaining a feeling of control over a decision‐making process, again even if that control is illusory, enhances one's sense that one has been treated fairly. In the words of Lind, Ruth Kanfer, and Christopher Earley, “[A]s long as there is an opportunity to express one's views and opinions before the decision is made, procedural fairness is enhanced” (1990: 952). Other research has suggested that such feelings of “process control” (Thibaut and Walker 1975) or “voice” (Folger 1977) can increase a subordinate worker's sense of satisfaction with a performance appraisal delivered by a manager (Korsgaard and Roberson 1995) and even enhance one's view of oneself (Rind and Kipnis 1999).

The present study was designed to explore the role of such findings in a negotiation context wherein parties first tried to influence their counterpart and then decided whether to accept or reject a concession offered by the other party. As such, the study differed from related research in that we were interested in a negotiation outcome, that is, whether the negotiators succeed in reaching agreement or not (cf. Heuer and Penrod 1986), rather than in their perceptions of fairness or control alone. In addition, in contrast to the strategy adopted in many investigations of procedural justice, we employed an experimental rather than a correlational approach (cf. Blader and Tyler 2003) in which we manipulated participants’ feelings of control over the process and outcome in the context of a realistic negotiation scenario (as opposed to a hypothetical role‐play) about an issue of genuine concern to them.

Specifically, we hypothesized that feedback from a negotiation counterpart (an experimental confederate) suggesting that participants had exerted some control over the content of the proposal with which they had been presented would lead them to view the negotiation process, the other party, and perhaps the content of the proposal itself in a more positive light. Any or all of these factors, we further hypothesized, would lead to an increase in agreement rates. In short, we were interested in whether acknowledgment by one side of the other side's role in winning a particular concession, even if that supposed influence was illusory, would facilitate agreement.

In this study we also explored several specific mechanisms by which agreement could be achieved through provision of this type of “acknowledgment.” First, recipients of a compromise proposal who receive such acknowledgment may view that compromise as more congruent with their expressed interests and priorities or simply as “better” for them. Second, such acknowledgment may lead the recipient of the proposal to feel that it represented a greater concession in terms of their counterpart's interests, and therefore that some reciprocal compromise on their own part is more called for. Alternatively, or perhaps additionally, the relevant acknowledgment may simply lead the recipients of a proposal to experience heightened positive sentiments toward their negotiation counterparts and thus to be more motivated to achieve a mutually satisfactory negotiation outcome rather than remain deadlocked.

The design of this study involved a negotiation between an undergraduate student participant who had indicated strong support for the legalization of marijuana on a prescreening survey and an undergraduate confederate who played the role of an anti‐legalization opponent — an opponent who nevertheless ultimately offered a compromise concession. In particular, the confederate presented negotiation participants either with an explicit and direct acknowledgment that the compromise about to be offered to them was a response to their expressed views or, alternatively, with a statement that the compromise had been formulated prior to the relevant negotiation. Although this procedure precluded the opportunity to study a genuine interaction between participants, it offered several important advantages in terms of experimental control. Most importantly, we were able to observe responses to a compromise proposal whose initial attractiveness to the participants remained constant across experimental conditions (and, as we shall detail, to create a context in which failure to compromise would be disadvantageous in terms of the goals and priorities of the negotiator). In the course of the study, we were also able to measure a host of evaluative and interpersonal assessments by the participants in those different conditions, thus allowing us to test a range of alternative explanations for any observed differences in the participants’ willingness or unwillingness to accept the relevant compromise.

Supporters of the legalization of marijuana took part in a dyadic negotiation with an anti‐legalization confederate of the same sex. (See “Participants” in the next section.) Negotiating pairs first discussed their opposing views and then were given the opportunity to make proposals to each other in an effort to reach consensus on a single plan for the limited legalization of marijuana.

In one condition, the confederate explicitly linked the content of his or her proposal to the views that the confederate's counterpart had expressed during the preceding discussion. In a second condition, the confederate essentially made an identical proposal, but with no such linkage or acknowledgment of the participant's views. In fact, the confederate in this “no acknowledgment” condition made it clear that the content of the proposal he or she was putting forward had been determined before hearing the participant's views. Participants in these two experimental conditions then had the chance to accept or reject the confederate's proposal and, if they chose, to make a counterproposal.

In the context of a mass‐administered prescreening questionnaire, undergraduates in an introductory psychology class were asked to indicate their level of “support for the legalization of ‘soft drugs’ (such as marijuana) in the U.S.” using a 7‐point scale (1 = strongly against; 4 = neutral; 7 = strongly in favor). Respondents who had indicated a strong pro‐legalization stance (by circling either 6 or 7) were identified and invited to participate in the study in partial fulfillment of a course requirement. Ultimately, thirty‐nine participants (twenty‐three men and sixteen women) completed the relevant procedures and were included in our statistical analyses.

On the day before they came to the laboratory, we informed participants that they would be taking part in a “negotiation exercise” in which they would be asked to consider proposals for the legalization of marijuana in California. They were further told that they were free to bring to the study any proposal of their own, but we also stated that this option (which was relevant to our main independent variable manipulation) was strictly optional.

Upon entering the laboratory, the participant was informed by the male experimenter that he was waiting for one more person to arrive before beginning the study. Presently, a same‐sex confederate, blind to experimental condition (see succeeding discussions) and playing the role of this “other participant” (whom, in the interest of clarity, we refer to as “she” or “her,” in contrast to our male experimenter) entered the room, and the study commenced.¹ The experimenter proceeded to explain that he was studying negotiation and that they would be taking part in a negotiation exercise using various materials that he had developed.²

In the next phase of the study, participants were reminded that California was currently considering the question of whether to legalize certain “soft drugs” such as marijuana (which, in fact, was true at the time the study was conducted). They were further told that a local political polling firm was interested in students’ reactions to legalization proposals and had agreed to use our participants’ responses in compiling data for recommendations to the state legislature, but “only in cases in which the two (negotiators) were able to agree on a single proposal.” At this point, the experimenter explained that he had brought them together because, on an earlier survey, they had expressed differing beliefs and attitudes with respect to the legalization of marijuana. He first asked that the person who had indicated general opposition to legalization on the earlier survey to identify himself or herself (at which point the confederate did so), and then asked the other participant to confirm that he or she had indeed previously expressed support for such legalization (which all participants did).

The experimenter informed both members of the dyad that they would be given five minutes to discuss their views on marijuana legalization with each other. He recommended to them that they use the time to explain any moral basis for their beliefs, relate personal experiences that had shaped their beliefs, or mention anything else that they felt was relevant to their positions. The experimenter left the room, and the discussion began. Adhering to a general script designed by the researchers, the confederate initiated the discussion by admitting that although she was generally opposed to the legalization of marijuana, she did not know a great deal about the issue. She then attempted to elicit the participant's views in a nonconfrontational manner. She listened attentively to the other participant, and if a response was called for, she simply stated that she did not see a compelling need to legalize marijuana. If pressed for further reasons underlying her position, she maintained that she did not want marijuana to become abused to the extent that alcohol was abused. Her demeanor remained calm, polite, and receptive but unyielding.

After five minutes had elapsed, the experimenter returned and invited the parties to begin the active negotiation phase of the study. Placing a sheet of paper containing ten specific sample proposals for the legalization of marijuana before them, he noted that “some people have found these useful when considering a possible proposal.” At that time he also indicated that “as I mentioned to each of you on the phone, you are also free to make any proposal that you might already have come up with and brought to the study today.” Nine of the stipulated proposals called for the limited legalization of marijuana coupled with a significant concession to legalization opponents (e.g., stiffer penalties for “hard” drug use or a quick end to legalization if evidence indicated that marijuana use had increased). The tenth proposal simply called for the legalization of marijuana use by all individuals. Participants were then asked to rate the attractiveness of the ten proposals on a separate sheet using a 7‐point rating, with anchors at 1 (not attractive at all), 4 (somewhat attractive), and 7 (extremely attractive).

As soon as the participant had completed his or her rating sheet, the experimenter discreetly observed the relevant responses to determine which proposal(s) had been rated a 4 (i.e., somewhat attractive). He then randomly selected one such proposal for use in the next phase of the study³ and used a prearranged code (which involved affixing a “subject I.D. number,” ostensibly accidentally omitted by the confederate herself, to her rating sheet) to covertly signal that choice to the confederate. Participants next were told that there would be only one “round” of negotiation, during which one person would have the opportunity to make a proposal and the other person would be entitled to accept or reject that proposal and, if he or she wished, to make a counterproposal.

The experimenter then turned to the confederate and remarked that she had previously “mentioned over the phone” that she had a proposal that she was willing to make. Turning to the participant, he asked if he or she similarly had come into the study with a proposal in mind. When, as was always the case, the participant indicated that the answer was “no,” the experimenter reassured the participant that it was “okay” and that no such proposal had been required of them. But he indicated that he wanted to give the “other participant” (i.e., the confederate) a chance to offer her previously prepared proposal. At that point, the confederate, utilizing the information that the experimenter had covertly signaled, offered a proposal that the participant had earlier rated as “4,” that is, “somewhat attractive.” In putting forward that proposal, however, she used one of two possible scripts, also signaled by the information conveyed by the experimenter, thereby achieving the independent variable manipulation featured in our study.

In the acknowledgment condition, which had a sample size of nineteen, the confederate told the participant that although she had come to the experiment with a previously prepared proposal, she now “in light of what you [i.e., the participant] said during the discussion” was going to “make a different proposal”— one, she noted, that was “actually quite similar to one of the proposals [they] had just rated.” In the no acknowledgment condition, which had a sample size of twenty, the confederate did not link that proposal to the content of the foregoing discussion. On the contrary, she indicated that her proposal, which was “actually quite similar to one of the proposals (they) had just rated,” was in fact the same one to which she had alluded in her prior phone conversation with the experimenter.

After the confederate had offered the relevant proposal, the participant was asked to respond formally to the offer by completing a rating form that asked several questions including, “How attractive is the proposal?”; “How large or significant a concession is this proposal on the part of the other side?”; and “How likely is it that you will accept this proposal?” Each question was followed by a 7‐point rating scale (1 = not at all; 4 = somewhat; 7 = extremely). A final item asked participants simply to circle the option that indicated whether they accepted or rejected the proposal, which would be announced to their counterpart (and, in the case of acceptance, allegedly submitted to the polling firm). The experimenter added that if the participant opted to reject the confederate's offer, the participant would then be free to offer a counterproposal.⁴

Upon the conclusion of the negotiation phase of the study, the experimenter gave the participants a post‐negotiation survey and then ushered the confederate out of the laboratory (ostensibly to write her answers privately in an adjoining lab room). The items on the questionnaire probed participants’ responses both to the proposal offered by the confederate and to the confederate herself, asking to what extent they felt that they had influenced the content of the proposal that had been made to them and to what extent it had taken their position into account. Seven‐point response scales anchored at 1 (not at all) and 7 (a great deal) accompanied both questions. Additional items asked participants to rate how similar they were to the confederate, how trustworthy they found her, how effective and sincere a negotiator she had been, and how much they liked her (again using appropriately anchored 7‐point scales). Participants were further asked the extent to which they thought their partner in the negotiation had been “only interested in what was best for him/her as opposed to being interested in satisfying you” (with the relevant response scale anchored at 1 = partner entirely self‐interested and 7 = partner entirely interested in satisfying me). Upon completing the questionnaire, the experimenter fully debriefed all participants and thanked them for their efforts.

Analysis of participants’ initial positions on marijuana legalization revealed an unpredicted significant difference between those who had been assigned to the “acknowledgment condition” and those who had been assigned to the “no acknowledgment condition.” Specifically, acknowledgment condition participants reported stronger support for legalization (mean = 6.74) than did no acknowledgement participants (mean = 6.25).⁵ Accordingly, participants’ initial position was included as a factor in all subsequent analyses.

One measure on the post‐negotiation questionnaire dealt with participants’ perceptions of the influence they had exerted on the concession‐making behavior of their negotiation counterpart (see Figure One). Because the logic of our conceptual analysis initially hinged on the assumption that the relevant acknowledgment manipulation would influence participants’ perceptions of personal influence and efficacy, this measure essentially constituted a manipulation check. A second measure probed participants’ sense that the confederate's proposal had actually taken their position into account. On the item that dealt directly with feelings about having influenced the content of the proposal they received, as anticipated, participants in the acknowledgment condition reported greater influence (mean =  4.26) than did participants in the no acknowledgment condition (mean = 3.25).⁶ On the measure that asked participants to assess the extent to which the other party had taken their position into account, however, the relevant difference in means for the acknowledgment and no acknowledgment conditions (mean = 5.05 and mean = 4.65, respectively) did not reach conventional significance levels (see Figure One).⁷ In other words, the acknowledgment condition participants’ sense that they somehow had influenced the confederate was stronger than their sense that the confederate's proposal had actually taken their position into consideration. These results thus suggest that although they felt some influence over the concession‐making process, these participants were not deceived into believing that a proposal of modest attractiveness completely addressed their concerns.

We then performed similar analyses on all assessment measures. The results indicated that participants in the acknowledgment condition rated the size of the confederate's concession to be somewhat larger (mean = 4.05) than did participants in the no acknowledgment condition (mean = 3.45; see Figure One).⁸ Of interest, however, the two groups did not differ significantly when it came to rating the attractiveness of the proposal (see Figure One).⁹ The relevant acknowledgment from the confederate thus seemed to heighten slightly participants’ conviction that the confederate was making a significant concession relative to her own previous position, even though the “revised” proposal did not in reality constitute an especially attractive one in terms of the participant's particular views and priorities. As confirmed by the relevant ratings, a proposal that participants had just rated as moderately attractive did not suddenly become highly attractive because of the relevant acknowledgment by the confederate.

Upon initially receiving the confederate's proposal, participants in the acknowledgment condition expressed a slightly greater likelihood that they would accept it than did participants who were denied such an acknowledgment (mean = 4.84 versus mean = 4.10, respectively; see Figure One).¹⁰ More importantly, when called upon to make their actual decision, nearly two‐thirds (63 percent) of the participants in the acknowledgment condition accepted the confederate's proposal, whereas less than half (40 percent) of the no acknowledgment condition participants accepted it (see Figure Two).¹¹

The final questionnaire administered to our participants dealt with their perceptions of the confederate who had offered them the concession. Examination of the pertinent items reveals that participants in the acknowledgment condition found her to be more “interested in satisfying them than satisfying herself” (mean = 4.21) than did participants in the no acknowledgment condition (mean = 3.70).¹² The acknowledgment condition participants also expressed greater liking for her (means = 5.68 versus 5.10, respectively; see Figure One).¹³ On two other measures, one assessing the confederate's trustworthiness and one assessing her effectiveness as a negotiator, the acknowledgment condition participants similarly rated the confederate more highly than did the no acknowledgment participants, although neither of these differences reached conventional levels of significance.¹⁴ Finally, participants in the two conditions did not differ significantly in how sincere they found the confederate to be¹⁵ or in how similar to themselves they found her to be.¹⁶

The acknowledgment manipulation in our study exerted a significant (or near significant) impact both on rates of acceptance of the confederate's proposal and on three sets of possible mediator variables — that is, perception of personal influence or efficacy, perception of the magnitude of the concession that the confederate had made, and feeling of positive interpersonal sentiments regarding the confederate who had made that concession. Relevant analyses revealed that the final variable, interpersonal sentiments, partially mediated the effect of acknowledgment on acceptance rates whereas the other two variables did not.¹⁷

Our results indicate that acknowledging the other side's expressed position in a negotiation, or more particularly, stating that one's concession has taken that position into account, facilitated acceptance of the relevant concession offer and, consequently, agreement. Such acknowledgment bolstered the recipient's belief that he or she had influenced the content of the proposal offered and that the concession in question had been a significant one.

It is worth reiterating that in the present study, the recipients of the proposal did not rate the proposal as more attractive when it was accompanied by the relevant acknowledgment, nor for that matter did they feel the concession maker had actually been more successful in taking their position into account. Of course, a significant change in ratings of attractiveness (or the perception that the proposal perfectly addressed the participant's concerns) would have been surprising in light of the fact that the proposal was, in all cases, one that the participant only moments before had rated as “somewhat attractive.” This set of ratings thus serves to underscore the nature of our main finding: we observed greater acceptance of a concessionary proposal deemed only modestly attractive when it was accompanied by a claim that its content had been altered to take into account the expressed interests of the recipient than when it was accompanied by implicit denial of such influence.

Post‐agreement assessments indicated that the relevant acknowledgment manipulation led to more positive feelings toward the person offering the concession. Moreover, mediational analysis suggested that the increased settlement rate in the acknowledgment condition was due more to this increase in the participant's positive feelings toward the confederate rather than to a change in his or her perceptions of personal efficacy or of the magnitude of the relevant concession. But the fact that the liking measure was taken after the relevant offer was accepted or rejected (in addition to the obvious limitations of any such correlational analysis) obliges us to be cautious in our speculations about causal mediation. Moreover, enhancement of positive sentiments may not facilitate agreement in all domains, particularly those in which the interpersonal reputations of the parties are less “in play.”

In essence, however, our results suggest that participants gave the confederate “credit” for trying to take their position into account, even though they felt she ultimately had been less than successful in that attempt. They then reciprocated that gesture by accepting a proposal that they otherwise would have been reluctant to entertain and subsequently expressed greater positive feelings for the confederate. The question of mediation and causal direction, of course, remains somewhat unclear. In particular, evidence that participants tended to show greater acceptance of the proposal from someone they came to like does not rule out the separate (and perhaps additional) tendency that participants came to like that individual as a consequence of having accepted his or her compromise.¹⁸

In interpreting the results of this study, it is important to recognize that the confederate in both conditions made a genuine concession by offering a compromise proposal that included at least limited legalization of marijuana. Coming from an individual who obviously opposed the legalization of drugs, such a concession (regardless of whether or not it was accompanied by specific acknowledgment of the participant's position) could have been perceived as a significant attempt to accommodate the views of the pro‐legalization advocates. This was especially true given that the mere possession of marijuana was illegal at the time and place in which the study was conducted. Nevertheless, the majority of participants in the no acknowledgment condition failed to reciprocate that concession by accepting the offer. Thus, reciprocity considerations (cf. Cialdini 1985) alone do not constitute a sufficient explanation for our results. Only when accompanied by the explicit statement that the concession in question was being made in response to the expressed views of the participant with whom the confederate was negotiating did the majority accept it. In the absence of that personal acknowledgment, and the positive sentiments it evidently evoked regarding the party offering it, the majority opted to stick with a status quo that was inferior to the offer on the table. In contrast to results obtained by Benton, Kelley, and Liebling (1972), the mere act of seeming to moderate one's position was not sufficient to convince the majority of participants to accept a compromise proposal. Instead, it took an explicit (albeit essentially untruthful) verbal acknowledgment of participants’ role in prompting that concession to convince most of them to accept it.

Similarly, although the confederate in the acknowledgment condition appeared to abandon a previously held proposal (i.e., the one she had “brought into the study”) in favor of one that might have been perceived by the participant as better addressing his or her needs, there is scant evidence that such a move resulted in the increased acceptance rates shown by participants in this condition. Indeed, research on so‐called reactive devaluation suggests that all else being equal, concessionary proposals that are withheld are likely to be preferred to those actually offered (see Ross and Ward 1995). At any rate, as stated earlier, the perceived magnitude of the concession did not appear to mediate the effect of acknowledgment on acceptance rates.

Finally, it was not the case that merely providing participants a “voice” in the process was sufficient to produce large‐scale agreement. Simply being permitted to air one's position in front of one's adversary did not constitute an effective tool in convincing the majority of participants to accept a compromise proposal, although it may well be the case that such a strategy is generally more effective than not permitting such an opportunity (Thibaut and Walker 1975; Lind, Kanfer, and Earley 1990). What proved effective in the present study was additional explicit acknowledgment by the person making the concession that the participant's relevant exercise of his or her voice had played a role in winning that concession.

Future research will help determine the degree to which a relevant acknowledgment must be made explicit in order to facilitate agreement. Additional studies may also shed light on whether acknowledgments must link a concession to the expressed views of the other party in order to be effective or whether it is sufficient for a potential concession offerer simply to indicate that one has “heard” the other side (see Johnson 1971).

Common wisdom dictates that a negotiator's success in securing an agreement depends largely on the attractiveness of the offer that is put on the table — a quality dictated largely by how closely it matches the other side's interests and priorities. The results reported in this article, however, suggest that an alternative path to success may be available: if one wants to increase the chances that an adversary will accept a compromise offer, then the key may lie less in enhancing the attractiveness of the relevant offer than in lowering the other side's threshold for acceptance. Our present findings suggest that this can be accomplished through explicit acknowledgment that one has attempted to accommodate the expressed position and interests of the other side.