Abstract
The current research shows that having no power can be better than having a little power. Negotiators prefer having some power (weak negotiation alternatives) to having no power (no alternatives). We challenge this belief that having any alternative is beneficial by demonstrating that weak alternatives create low anchors that reduce the value of first offers. In contrast, having no alternatives is liberating because there is no anchor to weigh down first offers. In our experiments, negotiators with no alternatives felt less powerful but made higher first offers and secured superior outcomes compared with negotiators who had weak alternatives. We established the role of anchoring through mediation by first offers and through moderation by showing that weak alternatives no longer led to worse outcomes when negotiators focused on a countervailing anchor or when negotiators faced an opponent with a strong alternative. These results demonstrate that anchors can have larger effects than feelings of power. Absolute powerlessness can be psychologically liberating.
When the prolific sports agent Leigh Steinberg signed rookie Steve Bartkowski as his first client in 1975, he soon learned that negotiating a contract in the National Football League would be a challenge because Bartkowski had no alternatives. Steinberg, the inspiration behind the movie Jerry Maguire, recently recalled that his client “would be forced to accept the team’s best offer. He could play quarterback for the Atlanta Falcons or play for no one” (Steinberg & Arkush, 2014, p. 43). Despite being virtually powerless in the negotiations, Steinberg made a bold move when he made the first official offer of $750,000: “We are asking for the largest amount of money any football player has ever been given” (p. 46). The Atlanta Falcons were outraged by Steinberg’s exorbitant request but eventually agreed to sign Bartkowski for $600,000, the most a rookie had received in the league’s history.
This example illustrates the surprising fact that negotiators who lack viable alternatives can achieve profitable agreements and stands in stark contrast to the advice that researchers and practitioners give to negotiators: to always secure an alternative (Malhotra & Bazerman, 2007; Thompson, 2011). It is widely recommended that negotiators establish alternatives because doing so gives them the power to extract more concessions from their opponents. Indeed, experimental research demonstrates that negotiators with better alternatives end up with superior outcomes because they can demand more (Kim, Pinkley, & Fragale, 2005; Mannix & Neale, 1993).
We challenge the widely held assumption that alternatives are necessarily beneficial and propose that the advice to secure an alternative can paradoxically have detrimental consequences. Drawing on research on anchoring effects, we hypothesize that someone with no alternatives (no power) can secure better outcomes than someone with a weak alternative (little power) because weak alternatives serve as low anchors that reduce the value of first offers. Such a finding would imply that (a) having no power can be better than having a little power and (b) anchor values can matter more than feelings of power in predicting negotiated outcomes.
Alternatives in Negotiations
Securing alternatives appears to provide a well-established advantage. One’s BATNA, or Best Alternative To a Negotiated Agreement (Fisher & Ury, 1981), is tied to one’s power. The better the best alternative, the more power one has because one is less dependent on a particular negotiation to achieve one’s goals. Negotiators with alternatives have higher aspirations, make larger initial demands, and claim more value than negotiators with no alternatives (see Thompson, Wang, & Gunia, 2010, for a review).
To test whether negotiators also prefer to have an alternative rather than no alternative, we asked 101 Mechanical Turk participants whether they preferred negotiating a job offer with a relatively unattractive alternative or with no alternative at all (see the Supplemental Material available online). An overwhelming majority (92.08%) chose to have a weak BATNA over no BATNA, preferring to have a little power rather than no power (see Handgraaf, Van Dijk, Vermunt, Wilke, & De Dreu, 2008, for similar findings in ultimatum games).
In the current research, we examined whether having a little power is actually better than having no power. Prior research has not answered this question because the comparisons have been between negotiators with attractive and unattractive BATNAs (Kim & Fragale, 2005) or between negotiators with attractive BATNAs and with no BATNA (Magee, Galinsky, & Gruenfeld, 2007; Van Kleef, De Dreu, Pietroni, & Manstead, 2006). As a result, it is not known how having unattractive alternatives, in comparison with having no alternatives, influences negotiators.
Anchoring and the Liberating Effect of Having No Alternatives
Although alternatives make negotiators feel powerful, the value of an alternative also serves as a salient anchor. We propose that the anchoring value of BATNAs can matter more than feelings of power in predicting negotiated outcomes.
Anchoring constitutes the assimilation of a judgment to a relevant or irrelevant numeric value (Tversky & Kahneman, 1974). Anchors have robust effects on economic outcomes and judicial verdicts (Englich, Mussweiler, & Strack, 2006; Northcraft & Neale, 1987). BATNAs can function as dominant anchor points. For instance, a BATNA determines a negotiator’s reservation price (the bottom line), which has a strong impact on the final agreement (Blount, Thomas-Hunt, & Neale, 1996).
We propose that negotiators are likely to use their BATNAs to determine the size of their first offers. Thus, weak alternatives may serve as low anchors that limit the size of first offers. Recall the football negotiation with which we opened this article: If Bartkowski had had an alternative offer of only $100,000, his offer to the Atlanta Falcons might have been much lower than the $750,000 he proposed. In contrast, not having an alternative might protect a negotiator from being anchored on a low value. Accordingly, we predicted that negotiators without a BATNA would make higher first offers than those with a weak BATNA because there would be no anchor to weigh their first offer down.
The anchoring effect of BATNAs on first offers is important because the size of first offers has a strong impact on final agreements (Galinsky & Mussweiler, 2001; Gunia, Swaab, Sivanathan, & Galinsky, 2013; Sinaceur, Maddux, Vasiljevic, Nückel, & Galinsky, 2013); aggressive first offers produce larger profits (Galinsky, Ku, & Mussweiler, 2009). We predicted that negotiators without alternatives would achieve more profitable agreements than negotiators with unattractive alternatives because without an anchor weighing them down, negotiators without alternatives would make more aggressive first offers.
Overview of the Experiments
We hypothesized an ironic effect of having no power: Negotiators with no BATNA would feel less powerful but make more aggressive first offers than those with weak BATNAs. That is, we predicted that the anchor value of alternatives would matter more than the feelings of power they induced. In our first experiment, we demonstrated that BATNAs influence first offers (Experiment 1). We then established the critical role of anchoring processes in the influence of BATNAs on final agreements by demonstrating mediation by first offers (Experiments 2 and 3) and moderation by whether negotiators focused on an alternative anchor (Experiment 3) or by the strength of an opponent’s BATNA (Experiment 4). All experiments were confirmatory studies. We report all conditions and excluded observations. Sample size was determined in advance.
The present studies make a number of important theoretical contributions. First, they demonstrate that, contrary to lay beliefs, having no power can be better than having a little power: Having no alternatives can psychologically liberate negotiators from the constraints that low anchors impose on first offers. Second, our work shows that anchoring effects can matter more than power in negotiations. Third, our finding that powerlessness can lead negotiators to demand more from their opponents extends prior research which found that people without any power are treated more sympathetically and generously than those with little power (Handgraaf et al., 2008).
Experiment 1: BATNAs Influence First Offers
In our first experiment, which comprised two separate studies, we manipulated whether negotiators had no, weak, or strong alternatives. We predicted that negotiators without alternatives would feel less powerful but make higher first offers than those with weak alternatives.
Experiment 1a
Participants and design
Three hundred five individuals (mean age = 30.50 years, SD = 9.27; 37.7% female) were recruited from Amazon’s Mechanical Turk in exchange for $0.50 and were randomly assigned to a no-, weak-, or strong-BATNA condition.
We excluded participants who had duplicate IP (Internet protocol) addresses or who responded incorrectly to an attention check. We also excluded participants who made first offers with extreme values (i.e., > 3 SD from the mean; McClelland, 2000). Seventeen participants were excluded for meeting one or more of these criteria, which left a final sample of 288.
Procedure
Participants were told to imagine that they had recently purchased a digital music player and therefore wanted to sell one of their old CDs. They were also given information about the quality of the CD (used but in reasonably good condition, with no scratches and an intact case). Then, they were informed that a potential buyer asked them for an initial offer. In the no-BATNA condition, participants were told: “Nobody else has offered you money for the CD. Thus, if you can’t reach an agreement in the current negotiation, you won’t get any money for the CD.” In the weak-BATNA condition, participants read: “Another buyer has offered you $2 for the CD. Thus, if you can’t reach an agreement in the current negotiation, you will get $2 for the CD.” Instructions were identical in the strong-BATNA condition, but the alternative offer was $8 instead of $2.
Immediately after the BATNA manipulation, participants were instructed to make the first offer, which was our key dependent measure. Then, the negotiation was terminated, and participants were asked to indicate the extent to which they felt powerful (1 = powerless, 7 = powerful), in control (1 = no control, 7 = in control), strong (1 = weak, 7 = strong), and confident (1 = unconfident, 7 = confident). Responses to these four items were averaged to create a single measure of perceived power (α = .92). Finally, participants completed an attention check (Oppenheimer, Meyvis, & Davidenko, 2009) and provided demographic information.
Results
Perceived power
We predicted that no-BATNA participants would feel the least powerful. That is exactly what we found (see Fig. 1). Negotiators with no BATNA (M = 4.80, SD = 1.03, 95% confidence interval, or CI = [4.59, 5.01]) felt less powerful than negotiators in both other conditions, ts(285) > 3.05, ps < .003, ds > 0.42. In addition, those with a weak BATNA (M = 5.25, SD = 1.09, 95% CI = [5.03, 5.47]) felt less powerful than those with a strong BATNA (M = 5.75, SD = 0.96, 95% CI = [5.55, 5.95]), t(285) = 3.34, p < .001, d = 0.49.
Results from Experiment 1a: mean first-offer amount (bars) and perceived power (lines) as a function of condition. Error bars indicate ±1 SEM. BATNA = best alternative to a negotiated agreement.
First offers
We also predicted that participants with a weak BATNA would make the lowest first offers, and we found that as well (see Fig. 1). Negotiators with a weak BATNA (M = $4.57, SD = 1.74, 95% CI = [4.33, 4.92]) made lower offers than negotiators in both other conditions, ts(285) > 6.18, ps < .001, ds > 0.76. Negotiators with a strong BATNA (M = $11.02, SD = 1.90, 95% CI = [10.63, 11.42]) made higher first offers than negotiators with no BATNA (M = $7.61, SD = 5.33, 95% CI = [6.54, 8.68]), t(285) = 6.83, p < .001, d = 0.85.
Thus, even though negotiators without alternatives felt less powerful than negotiators with unattractive alternatives, they made higher first offers.
Experiment 1b
In Experiment 1b, we used the same design as in Experiment 1a except that we included an ambiguous-BATNA condition to test whether the framing of the no-BATNA condition mattered. We also provided participants with several reference points to better correspond with the fact that negotiators have various sources of information available when making offers (Blount et al., 1996).
Participants and design
We recruited 204 individuals (mean age = 30.89 years, SD = 10.12; 30.9% female) from Mechanical Turk in exchange for $0.70 and randomly assigned them to the ambiguous-, no-, weak-, and strong-BATNA conditions. Nineteen participants were excluded using the same criteria as in Experiment 1a.
Procedure
Participants took the role of a job applicant and were asked to imagine that they had been invited for a job interview, that they had aced the interview, that the firm had extended a job offer, and that the only remaining issue was the salary. They were further asked to imagine that the human-resources representative had asked them for an initial offer for their desired salary. All participants were given salary estimates from a fictitious industry report indicating that for comparable positions, the salary range was from $30,000 to $80,000 and the mean was $60,000.
Participants in the ambiguous-BATNA condition were given no additional information. In the no-BATNA condition, participants read: “You do not have any alternative offers and you do not expect more to come.” Instructions in the weak-BATNA condition read: “A recruiter from a comparable firm has made you an offer with an overall value of $40,000.” Instructions in the strong-BATNA condition were identical except that the alternative offer was $70,000.
Immediately following the BATNA manipulation, participants were asked to make the first offer (our key dependent measure) and to complete the same four power items (α = .93) and attention check as in Experiment 1a.
Results
Perceived power
We predicted that no-BATNA participants would feel less powerful than participants with a BATNA. That is exactly what we found. Negotiators with no BATNA (M = 4.76, SD = 1.35, 95% CI = [4.37, 5.14]) felt less powerful than those with a weak BATNA (M = 5.30, SD = 1.07, 95% CI = [4.97, 5.62]), t(181) = 2.30, p = .023, d = 0.44, and less powerful than those with a strong BATNA (M = 5.69, SD = 0.94, 95% CI = [5.42, 5.96]), t(285) = 4.10, p < .001, d = 0.81 (see Fig. 2). Participants with a weak BATNA felt less powerful than those with a strong BATNA, but this difference was only marginally significant, t(285) = 1.69, p = .094, d = 0.40. Perceived power in the ambiguous-BATNA condition (M = 5.12, SD = 1.15, 95% CI = [4.76, 5.48]) differed from perceived power in the strong-BATNA condition, t(285) = 2.40, p = .018, d = 0.55, but did not differ from perceived power in the weak- and no-BATNA conditions (ps > .12).
Results from Experiment 1b: mean first-offer amount (bars) and perceived power (lines) as a function of condition. Error bars indicate ±1 SEM. BATNA = best alternative to a negotiated agreement.
First offers
We predicted that first offers would be the lowest in the weak-BATNA condition, and that is exactly what we found. Negotiators with a weak BATNA (M = $54,705, SD = 9,967, 95% CI = [51,674, 57,735]) made lower first offers than negotiators in all other conditions, ts(285) > 2.47, ps < .014, ds > 0.50 (see Fig. 2). Negotiators with a strong BATNA (M = $74,633, SD = 5,855, 95% CI = [72,951, 76,314]) made higher first offers than negotiators in the no-BATNA condition (M = $59,610, SD = 9,451, 95% CI = [56,924, 62,295]), t(285) = 7.79, p < .001, d = 1.91, and those in the ambiguous-BATNA condition (M = $60,762, SD = 12,470, 95% CI = [56,875, 64,648]), t(285) = 6.87, p < .001, d = 1.46. The no-BATNA and ambiguous-BATNA conditions did not differ, p = .56. Thus, the framing of the no-BATNA condition did not drive the observed effects.
Discussion
Experiment 1 demonstrated that negotiators without a BATNA made higher first offers than negotiators with a weak BATNA even though they felt less powerful. The anchor value of alternatives mattered more than the feelings of power they activated.
Experiment 2: Final Agreements and Mediation by First-Offer Amount
Experiment 2 involved an interactive negotiation. We predicted that negotiators without an alternative would reach more profitable agreements than those with a weak alternative and that this effect would be mediated by the size of the first offer. Also, to address the concern that participants in Experiment 1 may have taken the anchor provided by the experimenter as a signal of the true value, we used a negotiation item about which all participants had market knowledge. A confederate provided the alternative, to increase realism.
Participants and design
Participants were French students who demonstrated market knowledge of the negotiation item, a Starbucks logo mug. Potential participants were invited to participate in an online screening survey that ostensibly tested their Internet skills. The survey required them to search the price of the mug at the Starbucks online store. To disguise the purpose of the survey, we included two additional products to be researched. Only the 100 students (mean age = 22.47 years, SD = 3.08; 62.0% female) who correctly identified the retail price of the mug (€8.90) were invited to participate in the laboratory study in exchange for €5. At the lab several days later, participants were assigned to the role of either the seller or the buyer. The 50 dyads were then randomly assigned to either a no-BATNA or a weak-BATNA condition.
Although all dyads reached an agreement, seven dyads were excluded from analyses because the seller or buyer (or both) failed to correctly recall the retail price of the mug upon completion of the study. An additional dyad was excluded because their reported first offer was more than 3 standard deviations from the mean. All analyses were conducted on the remaining 42 dyads.
Procedure
BATNA manipulation
Upon arrival at the laboratory, participants were seated in separate cubicles, where they read their individual instructions. Sellers were endowed with a Starbucks logo mug. Buyers were told that they were about to meet the seller of a mug; sellers were told that before they entered the face-to-face meeting with the buyer (Buyer B), they would receive a phone call from another buyer (Buyer A, played by a male confederate), which might give them some leverage in the negotiation with Buyer B. To make sure that the confederate role was credible, we invited at least 4 to 6 participants to the lab at the same time and informed them that several other participants were delayed and would arrive shortly. Sellers were called by the confederate while in their individual cubicles. The confederate randomly communicated one of two statements and then hung up immediately. In the no-BATNA condition, the confederate said: “Hi, I’m Buyer A, and I promised to call back about the coffee mug you are offering. I’m not interested in your mug, and I cannot make you an offer for the mug. Have a nice day.” In the low-BATNA condition, the confederate said: “Hi, I’m Buyer A, and I promised to call back about the coffee mug you are offering. I’m interested in your mug, and my offer for your mug is €1.50. I’m unable to pay more than €1.50. Have a nice day.”
Negotiation
Following the phone conversation with Buyer A (the confederate), each seller was accompanied by a research assistant to the cubicle of Buyer B (another participant), where the buyer and seller were seated at a table and negotiated face-to-face. The seller was then instructed to make the first offer and to negotiate with Buyer B until they reached an agreement over the sales price of the mug.
Measures
After the negotiations, sellers indicated how powerful they had felt immediately after the phone conversation with Buyer A, using the same four power items as in Experiment 1 (α = .82). Both buyers and sellers reported the seller’s first offer and the final agreement amount, which served as our key dependent measures. Finally, all participants were asked to report demographic information and to choose from a set of three options (€2.90, €8.90, €16.90) the retail price of the mug according to their research during the prelab survey.
Results
Perceived power
As we predicted, negotiators with no BATNA (M = 3.51, SD = 1.03, 95% CI = [3.02, 4.01]) felt less powerful than those with a weak BATNA (M = 4.32, SD = 1.26, 95% CI = [3.77, 4.86]), t(40) = 2.22, p = .032, d = 0.70 (see Fig. 3).
Results from Experiment 2: mean amount of the final agreement (bars) and perceived power (lines) as a function of condition. Error bars indicate ±1 SEM. BATNA = best alternative to a negotiated agreement.
First offers
Also as predicted, negotiators in the no-BATNA condition (M = €10.94, SD = 4.25, 95% CI = [8.89, 12.99]) made higher first offers than those in the weak-BATNA condition (M = €4.39, SD = 2.17, 95% CI = [3.45, 5.33]), t(40) = 6.46, p < .001, d = 1.94.
Final agreements
Similarly, we expected no-BATNA participants to negotiate more profitable agreements than weak-BATNA participants. Indeed, negotiators with no BATNA (M = €7.79, SD = 3.19, 95% CI = [6.26, 9.33]) negotiated better agreements than those with a weak BATNA (M = €3.73, SD = 1.82, 95% CI = [2.95, 4.52]), t(40) = 5.18, p < .001, d = 1.56 (see Fig. 3).
Mediation analysis
We predicted that sellers with no BATNA would negotiate more profitable agreements than those with a weak BATNA because they made higher first offers. In a process analysis using a bootstrapping procedure with 5,000 iterations (Hayes, 2013), we designated the seller’s BATNA as the independent variable (weak BATNA = 0; no BATNA = 1), the seller’s first offer as the mediator, and the amount of the final agreement as the dependent variable. As predicted, the amount of the first offer mediated the effect of the negotiator’s BATNA on the size of the final negotiated price, 95% CI = [3.21, 6.49] (see Fig. 4).
Results from Experiment 2: first-offer amount as a mediator of the relationship between BATNA level (BATNA = best alternative to a negotiated agreement) and the size of the final agreement. The values along the paths are unstandardized regression coefficients, with standard errors in parentheses. Coefficients to the right of the slashes are simultaneous regression coefficients. Asterisks indicate significant paths (**p < .001).
Discussion
Experiment 2 demonstrates that having no alternative leads to more profitable agreements than having a weak alternative and that first-offer amount mediates this relationship. This study also establishes external validity of the observed effects, as all negotiators had market knowledge of the negotiation item and the BATNA was provided by another person rather than the experimenter.
Experiment 3: Leveling the Playing Field
Experiment 3 also involved an interactive negotiation. In this experiment, we used moderation to further test whether anchoring drove the observed effects. Specifically, we manipulated whether negotiators focused on a countervailing anchor. Galinsky and Mussweiler (2001) demonstrated that focusing on the countervailing anchor of one’s target price eliminates the anchoring effect of the first offer. We predicted that negotiators would be less anchored on a weak alternative when they focused on their target price.
Participants and design
Two hundred eighty-six French students (mean age = 22.99 years, SD = 2.83; 57.0% female) were recruited to participate in a laboratory study in exchange for €5. The 143 dyads were randomly assigned to six conditions in a 3 (BATNA: no vs. weak vs. strong) × 2 (focus: BATNA vs. target price) between-subjects design.
Although all dyads reached an agreement, 2 dyads did not follow the instructions regarding who was to make the first offer, and 2 dyads reported first offers that fell more than 3 standard deviations from the mean. These four dyads were dropped from the analyses. Thus, all analyses were conducted on the remaining 139 dyads.
Procedure
Upon arrival at the laboratory, each dyad was seated in an individual room and received their role materials. The scenario involved an antique dealer (the seller) and a collector (the buyer) negotiating the sales price for a sugar bowl (Paulson, 2014). All sellers were informed that they had originally purchased the sugar bowl for another client who had been willing to pay €600 but ultimately refused to accept the bowl because it had a defect. Sellers were also told that the market value for sugar bowls varies widely and can range up to €1,000.
Manipulations
BATNAs were manipulated similarly to the way they were manipulated in Experiment 1. Because past research suggests that making the first offer is equally effective for sellers and buyers (Galinsky & Mussweiler, 2001; Gunia et al., 2013; Moran & Ritov, 2002), we always instructed sellers to move first. Sellers in the no-BATNA condition were told that no one else had made them an offer, that they should not expect any more offers to come, and that they would earn no money if they did not reach an agreement. Sellers in the weak-BATNA condition were told that another buyer had offered them €45 for the bowl and that they would earn this amount if they did not reach an agreement. The instructions in the strong-BATNA condition were identical except that the BATNA value was €450.
Following the BATNA manipulation, sellers received an additional page of information, titled “Important Seller Information,” that instructed them to focus on either their BATNA or their target price (Galinsky & Mussweiler, 2001). The BATNA-focus instructions read, “When preparing for your negotiation it is important to think about and focus on the alternatives that you have to this negotiated agreement. A clear understanding of these alternatives will assist you in preparing for the negotiation.” The target-price-focus instructions read, “When preparing your negotiation it is important to think about and focus on your target, the ideal price at which you could sell. A clear understanding of this price will assist you in preparing for the negotiation.”
Because both manipulations were directed at sellers, all buyers received identical instructions and were told that they should pay no more than €3,000 for the bowl.
Measures
After reading the role instructions but before the first offer was made, sellers were asked to complete the same four power measures as in Experiments 1 and 2 (α = .86). When the negotiation was completed, participants reported the value of the first offer made by the seller and the amount of the final agreement, which served as our key dependent measures. 1 Finally, all participants reported demographic information.
Results
Perceived power
Results for the BATNA-focus condition were consistent with the results in Experiments 1 and 2. When negotiators focused on their BATNA, those with no BATNA felt the least powerful, those with a weak BATNA felt more powerful, and those with a strong BATNA felt most powerful (see Table 1). There was a significant difference between the no- and strong-BATNA conditions, but the weak-BATNA condition did not differ from the other two. However, when negotiators focused on their target, there were no differences among the no-BATNA, weak-BATNA, and strong-BATNA conditions (ps > .41; see Table 1).
Perceived Power, First-Offer Amounts, and Final-Agreement Amounts in Experiment 3
Note: The table presents means, with standard deviations in parentheses and 95% confidence intervals in square brackets. Within a row, means with different subscripts differ significantly from each other, p < .05. BATNA = best alternative to a negotiated agreement.
First offers
We predicted that negotiators who had weak BATNAs and who focused on their BATNA would make lower first offers than negotiators in each of the other conditions. In contrast, we expected that negotiators who had weak BATNAs but focused on their target price would show the same aggressive first offers as participants in the no-BATNA condition. This is exactly what we found. First offers were significantly lower in the weak-BATNA/BATNA-focus condition than in all other conditions, all ts(133) > 2.48, all ps < .015, all ds > 0.98. However, when negotiators focused on their target price, the differences among the BATNA conditions were reduced, and negotiators in the weak-BATNA condition made significantly higher first offers when they focused on their target than when they focused on their BATNA (see Table 1).
Final agreements
We predicted that final agreements would show the same pattern as first offers, and this is what we found. Final agreements in the weak-BATNA/BATNA-focus condition were significantly different from final agreements in all other conditions, all ts(133) > 2.99, all ps < .004, all ds > 1.04. However, when participants focused on their target price, the differences among the BATNA conditions were reduced, and negotiators with a weak BATNA achieved more profitable agreements when they focused on their target price than when they focused on their BATNA (see Fig. 5 and Table 1).
Results from Experiment 3: mean amount of the final agreement as a function of condition. Error bars indicate ±1 SEM. BATNA = best alternative to a negotiated agreement.
Mediation analysis
We hypothesized that the size of the first offer would explain why negotiators with a weak BATNA reached lower outcomes than those with no BATNA when negotiators focused on their alternative, but not when they focused on their target price. To test for mediation, we carried out a process analysis using a bootstrapping procedure with 5,000 iterations (Hayes, 2013; Preacher, Rucker, & Hayes, 2007). As in Study 2, we designated the seller’s BATNA as the independent variable (weak BATNA = 0; no BATNA = 1; n = 97), the seller’s first offer as the mediator, and the amount of the final agreement as the dependent variable. As predicted, when negotiators focused on their BATNA, the amount of the first offer mediated the effect of BATNA condition on the final agreement, 95% CI = [58.06, 263.53]. However, when negotiators focused on their target price instead, no mediation was found, 95% CI = [−4.35, 159.71] (see Fig. 6).
Results from Experiment 3: first-offer amount as a mediator of the relationship between BATNA level (BATNA = best alternative to a negotiated agreement) and the size of the final agreement when negotiators focused on their BATNA (top panel) and their target price (bottom panel). The values along the paths are unstandardized regression coefficients, with standard errors in parentheses. Coefficients to the right of the slashes are simultaneous regression coefficients. Asterisks indicate significant paths (*p < .01, **p < .001).
We also tested an alternative mediation model recommended by Hayes and Preacher (2014), using a multicategorical predictor (−5 for the weak-BATNA/BATNA-focus condition, 1 for all other conditions). Again, first-offer amount mediated the causal effect of the predictor on the amount of the final agreement, 95% CI = [9.70, 44.05]. Thus, our mediation analyses support our theory that not having an alternative frees negotiators up to make higher first offers and, as a consequence, achieve more profitable agreements.
Discussion
No-BATNA negotiators achieved more profitable agreements than those with a weak BATNA. We established the role of anchoring in this effect using both mediation and moderation. The effect of BATNA condition on outcomes was mediated by first-offer size and did not emerge when negotiators focused on their target price.
Experiment 4: The Opponent’s Alternatives
We next tested whether the liberating effect of having no alternative depends on the perception of the opponent’s alternatives. The strength of an opponent’s alternatives is important because negotiators with stronger alternatives are seen as more powerful (Pinkley, 1995). As a result, no-BATNA negotiators may lower their expectations and first offers if they think that their opponent has a strong BATNA. In our final study, we manipulated the strength of the opponent’s BATNA to test whether it would moderate the previously observed effects.
Participants and design
Four hundred five individuals (mean age = 28.60 years, SD = 9.06; 35.8% female) recruited from Mechanical Turk participated in this experiment in exchange for $0.50. Participants were randomly assigned to a 2 (own BATNA: no vs. weak) × 4 (opponent’s BATNA: ambiguous vs. no vs. weak vs. strong) between-subjects design. Twenty-six participants were excluded using the same criteria as in Experiment 1, which resulted in 379 observations.
Procedure
We used a modified version of the task in Experiment 1a. Participants always took the role of the seller.
Manipulations
The manipulation of own BATNA included only the two conditions of interest. Participants in the no-own-BATNA condition were told that they had no alternatives, whereas those in the weak-own-BATNA condition were informed that their alternative was $2.
The manipulation of opponent’s BATNA included the four conditions in Experiment 1b. Participants in the ambiguous-opponent-BATNA condition were not given any information about their opponent’s alternatives, those in the no-opponent-BATNA condition were told that their opponent did not have any alternatives, those in the weak-opponent-BATNA condition were told that their opponent’s alternative was to pay $15, and those in the strong-opponent-BATNA condition were told that their opponent’s alternative was to pay $5.
Measures
Subsequently, negotiators were prompted to make their first offer, complete our four-item power scale (α = .92), and respond to an attention check and demographics questions.
Results
Perceived power
Results were consistent with those in Experiments 1 to 3. Negotiators with no BATNA felt less powerful than those with a weak BATNA irrespective of the opponent’s BATNA, ts(371) > 2.92, ps < .004, ds > 0.71 (see Table 2). A contrast analysis further revealed that negotiators generally felt less powerful in the strong-opponent-BATNA condition (M = 4.93, SD = 1.28) compared with all other opponent-BATNA conditions (M = 5.24, SD = 1.12), t(375) = 2.24, p = .025, d = 0.27.
Perceived Power and First-Offer Amounts in Experiment 4
Note: The table presents means, with standard deviations in parentheses and 95% confidence intervals in square brackets. For each measure, means with different subscripts differ significantly from each other, p < .05. BATNA = best alternative to a negotiated agreement.
First offers
We predicted that negotiators in the weak-own-BATNA conditions would make lower first offers than those in the no-own-BATNA conditions except when the opponent had a strong BATNA. This is exactly what we found (see Fig. 7 and Table 2): Negotiators with a weak BATNA made lower first offers than negotiators with no BATNA when their opponent’s BATNA was perceived to be weak, t(371) = 4.61, p < .001, d = 1.10; absent, t(371) = 5.50, p < .001, d = 0.80; or ambiguous, t(371) = 5.69, p < .001, d = 1.08. However, when the opponent’s BATNA was strong, the difference between the weak- and no-own-BATNA conditions was much smaller and no longer significant, t(371) = 1.56, p = .12, d = 1.10.
Results from Experiment 4: mean first-offer amount as a function of condition. Error bars indicate ±1 SEM. Asterisks indicate significant differences between conditions (**p < .001). BATNA = best alternative to a negotiated agreement.
Discussion
Experiment 4 demonstrated that the effect of having no BATNA on first offers remains robust regardless of opponents’ BATNAs except when negotiators perceive their opponents to have a strong BATNA: No-own-BATNA negotiators no longer asked for more than weak-own-BATNA negotiators when their opponent had a strong BATNA.
General Discussion
The present research demonstrated that having no power can be better than having a little power. We found that the absence of alternatives can paradoxically liberate negotiators to make more aggressive first offers and achieve superior outcomes.
BATNAs are both a source of power and anchor values. The current experiments showed that the anchoring effect was stronger than the power effect. The results for first-offer amounts were consistent with the idea that BATNAs serve as anchors; negotiators made more aggressive demands when alternatives were absent than when they were weak. Thus, the very same resource (BATNA) that enables individuals to achieve superior outcomes when it is strong is constraining if it is weak, and its complete absence is liberating.
We demonstrated the role of anchoring through mediation and moderation. First, the effects of having a weak rather than no BATNA on final agreements were consistently mediated by first-offer amount. Second, when negotiators with a weak BATNA focused their attention on a countervailing anchor value—their target price—first offers were no longer weak, and final outcomes were no longer worse.
Certain features of our paradigm provide exciting avenues for future research. First, we always instructed focal negotiators to make the first offer. Although past research suggests that negotiators without alternatives may be less likely to initiate the negotiation than are negotiators with strong alternatives (Magee et al., 2007, Experiment 3), future studies should test whether no-BATNA negotiators are less likely than weak-BATNA negotiators to move first, for instance, by using a less controlled setting in which negotiators are free to move first or wait for their opponent to initiate the negotiation (see the Supplemental Material available online). Second, although none of the dyads in Experiments 2 and 3 reached an impasse, follow-up studies could explore the conditions under which opponents could be offended and walk away from high first offers by no-BATNA negotiators (see Schweinsberg, Ku, Wang, & Pillutla, 2012). Third, BATNAs are especially consequential for negotiation outcomes when both parties are aware of them (Pinkley, 1995). Thus, future research could explore whether the findings from our final study extend to situations in which negotiators are informed about each other’s BATNAs. Fourth, although we focused on BATNAs as the source of power, other forces, such as the relative contributions each party brings to the bargaining table, can further influence negotiators’ objective and subjective feelings of power (Kim & Fragale, 2005; Kim et al., 2005). Thus, future research is needed to test whether our findings hold when power is not derived from one’s BATNA. Finally, another issue that future research might address is whether our findings extend to complex, multi-issue negotiations in which making the first offer can backfire (Loschelder, Swaab, Trötschel, & Galinsky, 2014; Pinkley, Neale, & Bennett, 1994).
Conclusion
The present research reinforces the idea that anchors have considerable weight in negotiations. Because alternatives not only offer power but also serve as anchors, a weak alternative can weigh down the size of a first offer and produce a worse outcome than having no alternatives would have. As in Leigh Steinberg’s negotiation, having no power can be a liberating experience.
Footnotes
Acknowledgements
Portions of this research were conducted at the INSEAD Social Science Research Centre. We wish to thank Liselott Petterson and Huong Ngo for their valuable assistance with translation and data collection, as well as Stefan Thau, Martin Schweinsberg, Christilene Du Plessis, and the INSEAD OB brownbag participants for their helpful comments and suggestions.
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.
Funding
This research was partially supported by a grant from the INSEAD Alumni Fund.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
