Getting (More of) What You Want

Negotiators often attempt to influence their counterparts by issuing threats or making promises. For example, they might threaten to walk away if their counterpart does not concede on an issue, or even threaten to take an action that is outside the negotiation; for example, unions typically threaten to strike, thereby imposing costs on management if they don’t concede. On the other hand, negotiators also may promise to take an action that is outside the temporal frame of your current negotiation, such as: “If you concede on this issue, I will make sure that everyone knows what a great product you sell.”

In this chapter, we focus on promises and threats of actions that occur after the negotiation has concluded as forms of influence. Thus, when the target of the threat or promise concedes, he does not know if the issuer will actually carry out the threat or honor the promise.

Obviously, to be effective at influencing the target, promises and threats have to be credible; that is the target must believe that, if he concedes, the issuer will make good on the promise—or carry out the threat, if he does not. Conversely, if the target believes that the threat or the promise is not credible (i.e., that it will not be carried out), then he should simply ignore that threat or promise. Of course, it also follows that issuers should only threaten or promise when those actions are perceived as credible by the target and thus likely to have the desired influence.

We’ll begin by discussing the differences and the similarities between threats and promises. Then we focus on what makes promises and threats credible. We start with the assumption that both issuer and target act reasonably (or in Thomas’s view, rationally). We will then expand our analysis to include the psychological perspective.

PROMISES VERSUS THREATS

Both threats and promises are used to influence your behavior. For example, airlines routinely promise their best customers access to seats with more legroom and upgrades to first class. Those promises are contingent on availability and, thus, not really enforceable—yet these benefits are effective in influencing customers’ behavior only if they believe that the airline will make good on that promise. It is the same in a negotiation; promises are useful to the extent that you believe that your counterpart will honor her promises.

Just as in the case of promises where your counterpart can offer attractive inducements, she may threaten some other action that is costly to you if you do not concede. However, from your perspective (as the target of the action), an obvious difference between threats and promises is that if carried out, threats impose costs on you while promises offer benefits.

Carrying out either a threat or a promise, however, is costly to the issuer. Indeed, although it seems natural to focus on the costs and benefits from the perspective of the target of these attempts to influence, as we discuss next, what matters most for the credibility of the threats and promises themselves are actually the costs from the perspective of the issuer. After all, from a purely rational perspective, once a threat or promise is issued, there are real costs associated both with carrying out the threat or promise and also with failing to carry it out. In the first instance, the issuer must actually bear the cost of carrying out the action; in the second instance, failing to carry out the action will negatively affect the reputation of the issuer.

A second difference between threats and promises is that promises can be made contractually binding (at which point they become part of the settlement), thereby increasing their credibility. For example, promises can be made more credible if they are in the forms of warranties, which are typically legally enforceable. In contrast, threats generally cannot be made legally binding. For example, one cannot sign a legally binding contract to end the negotiation if you do not concede. Thus, making threats credible is a more challenging task—but there are ways to make threats more binding to the issuer and, thus, more credible.

A third difference between threats and promises lies in the target’s actions. In the case of threats, the target has no incentive to induce the issuer to carry out the threat, if the issuer were to back down. However, in the case of promises, the target wants the issuer to honor the promise. Because a promise can be carried out at a future date, the target may issue its own threat, for example to publicize the issuer’s failure to keep the promise.

Although the previous distinctions mainly focused on the economic aspect, from a psychological perspective, there is another important distinction. Promises and threats differ in that promises are associated with gains while threats are associated with losses, and there is a big difference in how people respond to potential gains and how they respond to potential losses. Therefore, framing your influence attempt as a threat or promise affects the amount of risk that your counterpart may be willing to accept.

To illustrate this psychological effect, consider the classic dread-disease example:1

The United States is preparing for the outbreak of an unusual Asian disease that is expected to kill up to 600 people. Participants in this experiment were randomly assigned to two groups. Group I was presented with the following two alternative programs and asked to select the one they favored:

1. If Program A is adopted, 200 people will be saved.

2. If Program B is adopted, there is a one-third probably that all will be saved and a two-thirds probability that none will be saved.

When presented with these two options, 76 percent of Group I participants chose Program A, while 24 percent chose Program B. Thus, it seems that Group I participants valued the prospect of saving two hundred lives for certain more than they valued the risky prospect, even though both programs were expected to save the same number of people.

Participants in Group II were presented with the following options:

1. If Program a is adopted, 400 people will die.

2. If Program b is adopted, there is a one-third probability that no one will die and a two-thirds probability that all will die.

13 percent chose Program a, while 87 percent chose Program b. Thus Group II participants expressed a strong preference for the riskier Program b, rather than the certain Program a. The prospect of 400 people dying for certain was less attractive than a lottery of equal expected value for Group II participants.

Note that from a rational perspective, both groups saw identical alternatives. For example, if 200 of 600 will be saved (Group I), then 400 will die (Group II). Similarly, if all are saved (Group I), then none will die (Group II). So from a purely rational perspective, both sets of participants should have ranked the two proposals equally. But they did not!

What these results show is that framing a choice as a potential gain (saving lives) makes the certain option more attractive, while framing a choice as a potential loss (people will die) makes the risky option more attractive.

To translate this effect into the language of negotiation: the risky choice is to resist accepting a proposal, not agreeing and hoping for a better alternative in the future. That this future option may not materialize is the risk here. Framing a choice as a gain results in your being more risk averse and increasing the probability that you will agree to the proposal. The risk-averse choice is to accept what is currently being proposed—what is certain.

In the same way, threats focus on what the target has to lose and, thus, invoke a loss frame that encourages resistance. In contrast, a promise focuses on what the recipient of the promise can gain and, thus, frames the interaction as a potential gain and encourages acceptance of the proposal. Thus, the way you frame your proposals can significantly alter your counterpart’s response.

THE POWER OF THREATS AND THE ALLURE OF PROMISES

The uncertainty about whether a threat or promise will be carried out stems from the temporal separation of issuing the threat or promise and its potential effect on the target. Once a threat is issued, the target must choose to concede or not, not knowing whether the issuer will actually follow through.

If the target were not to concede, the issuer then must decide whether to carry out the threat. But here lies the dilemma: the target has not conceded. His failure to concede has already happened. The threat has proved ineffective. Will the threat be implemented?

So what is a target to do? Following our advice from Chapter 6 to look ahead and reason back, begin your analysis in reverse chronological order. Will the issuer carry out the threat or make good on the promise? Let’s first consider the one-time negotiation where there is little chance of future interaction. For example, you may negotiate a purchase while on vacation in a city where future interaction with the seller is unlikely. But even locally, some purchases are simply not repeated with sufficient frequency to make the chance of future interaction a relevant consideration for either party. For example, the car dealer may not attach much weight to the possibility of your purchasing future vehicles from him.

So the issuer of the threat or promise makes a calculation: Is there a reasonable risk that the target will make the issuer’s bad behavior public? If the parties do not expect to interact in the future, from a rational perspective the target is not likely to make the bad behavior of the issuer public by advertising that a threat was ignored or a promise not honored. The reason for this inaction is that advertising such behavior requires the target to engage in costly activity such as holding up a sign in front of the dealer’s showroom advertising that promises were not honored. That action is costly to the target and has little potential for offsetting economic benefit that would justify that effort.

To illustrate, consider the following scenario: A start-up is considering entering a market dominated by an incumbent firm. Anticipating the start-up’s entry into its market, the incumbent firm privately issues a threat to fight a price war if the start-up goes through with its plans.2

If the start-up does not enter the market, the present value of the incumbent firm’s profits is $300 million, and the present value of the start-up’s profits is $0 (since it does not conduct business). However, if the start-up were to enter the market, the incumbent firm can either accommodate (ignore the threat it issued and simply accept the presence of the start-up) or carry out its threat and wage a price war. If it were to accommodate, the present value of the profits to the incumbent firm and the start-up are $100 million each. In contrast, if the incumbent firm were to decide to fight a price war, the incumbent stands to lose $100 million and the start-up stands to lose $200 million.

If you were an advisor to the start-up, how seriously would you take the incumbent’s threat of a price war? At issue is the potential for your client to lose $200 million if the incumbent were to fight a price war and a $100 million profit if it did not.

In this scenario, the threat of a price war is not credible, and you should advise your client to proceed with its plans to enter the market. To arrive at this answer, consider first the profits and losses from the incumbent firm’s perspective. What are the alternatives available to the incumbent once your client entered the market? The incumbent firm will earn $100 million by accommodating or lose $100 million by fighting a price war. Hence, once your client entered the market, it is more advantageous for the incumbent firm to accommodate than to carry out its threat. Thus, if you believe that the incumbent firm will act rationally, the threat of a price war is not in the incumbent firm’s best interest once the start-up has entered the market and, hence, not credible.

So from the start-up’s perspective the options are (a) not to enter the market and earn zero or (b) to enter the market and earn $100 million. Obviously the best alternative is to ignore the threat (since it is not likely to be carried out) and enter the market.

Of course, the situation is more complex if there were multiple potential start-ups who could enter the incumbent’s market. In that case, the incumbent firm may issue the threat publicly and, if your client were to enter, fight a price war to establish a reputation of being tough in defending its market—with the primary goal of dissuading other potential entrants.

The situation is simpler when it involves promises, rather than threats. The issuer will compare the costs of making good on the promise to the reputational costs of reneging. If no future interaction with the target is expected, those reputational costs will arise only if the recipient decides to make the issuer’s bad behavior public. But is that likely to occur?

So long as the issuer believes that the recipient is not willing to incur the costs to advertise the issuer’s bad behavior, he will not honor his promise. But then, knowing that the issuer is unlikely to honor his promise, the recipient is not likely to be influenced by the promise of the promise. Without the opportunity for future interactions, the issuer will rationally renege on the promise, and the rational recipient will not be influenced by promises. Thus, when future interactions are not likely, threats and promises should typically not be issued; and, if issued, they are likely to have little influence on the target.

So in a rational world—and in a one-and-done situation—neither promises nor threats are going to be particularly effective in influencing the target. In turn, because threats and promises are not going to be effective forms of influencing the target, it is not rational to issue them in the first place.

But threats and promises are routinely issued in such situations, and they have influence on negotiators even when there is likely to be no future interaction. So what psychological factors make threats and promises effective in influencing the behavior of targets and honored by their issuers when rational actors would ignore them? If you were trying to predict the actual behavior of a target or an issuer, you will need to consider more than just a rational analysis.

THE PSYCHOLOGICAL ASPECTS OF THREATS AND PROMISES

What may surprise you is that implementing some threats may actually be a positive psychological experience. The term “schadenfreude” refers to the pleasure derived from seeing the misfortunes of others. In the context of threats, the issuer may actually enjoy implementing the threat. If you threaten to key your competitor’s car, particularly if his car is very nice, you might actually take pleasure from his misfortune (at your hands). While schadenfreude is admittedly a primitive urge, it is real and, as such, must be taken seriously. But even with the benefit that the issuer gets from schadenfreude, executing a threat is still costly. Therefore, the enjoyment of schadenfreude has to exceed the costs of carrying out the threat for the threat to be credible. As such, a threat that is of low cost to the issuer to implement (keying the target’s car) is much more credible than a threat that is costly for the issuer to implement. You have to take a threat to key your car much more seriously than you would take a threat to steal your car.

Psychological benefit can also be experienced when you publicize a wrong that you experienced or acknowledge exemplary behavior. Have you ever gone out of your way to do so, even when you were reasonably sure that you would never interact with that person or organization again? Why did you bother?

One reason may be your belief in a just world: that people get what they deserve.3 The psychological cost to you of letting bad behavior go unpunished or good behavior go unnoticed could exceed the economic cost of the resources (for example, your time) of taking action. What if a person gains benefits in exchange for a promised future behavior—and then reneges on her promise? Getting undeserved benefits may challenge your belief that the world is a just place. So you may be motivated to take action to alter the situation to reestablish the balance.4 You might expose the person’s behavior in social media, write reviews of their business, share the information with your friends or classmates, or picket the establishment to publicize your counterpart’s private failure. Take the case of many online sellers on eBay or Amazon. It might be very difficult to hold such multiple, virtual sellers to reasonable standards of honest and good behavior in their one-and-done transactions. To counteract this, sellers agree to be publicly rated by their customers. Thus, the cost to customers to advertise the bad behavior of sellers is minimal: a few keystrokes; and the potential damage to a seller’s reputation is large because the dissemination of this information is public and easily accessible by others. As a result, promises by sellers on eBay’s or Amazon’s networks (such as speedy delivery, merchandise as advertised, or hassle-free returns) are more credible because of the public feedback about the sellers’ reputations.

The extent to which you, as an individual, are willing to go out of your way to make public the bad (and good) behavior of your counterpart is associated with how strongly your beliefs in a just world have been violated, even when there is likely to be no future interaction. Of course, not all individuals are equally influenced by a just-world belief. When negotiating, it may be useful to convince your counterpart that you are a fierce believer in a just world and that behaviors—good or bad—will be made public. Then, concede in response to her promises only if you reasonably believe that you have convinced her of your sincerity.

Although the threat of making bad behavior public may deter your counterpart from behaving badly in the first place, the power of threats made in public is that they are more credible than threats made privately. The public nature of both threats and promises allows future partners to learn about the reputation of a counterpart. In essence, the issuer puts his global reputation (as opposed to the reputation with an individual counterpart) on the line for all to see. In turn, if that reputation is valuable, then the target must reasonably expect the issuer to follow through to preserve this valuable asset.

Even the public nature of threats, however, does not always result in costly threats being implemented. Take President Obama’s now infamous “red line” remark in 2012 regarding use of chemical weapons by Bashar al-Assad’s regime in Syria.5 President Obama publicly drew that red line. However, after it became clear that Assad’s regime did use chemical weapons on its own people, no action followed (unless one calls the request that Congress consider authorizing an action a decisive act—but most would not). As a result, President Obama’s reputation was tarnished. Not following through on this threat not only may have reduced his credibility with Assad but also may have encouraged others to question what other threats or promises he might also ignore. Thus, issuing of a threat or promise publicly increases its credibility precisely because it makes the costs of reneging so high.

While the public nature of a threat or promise is an important consideration in determining its credibility, other characteristics can affect how influential threats or promises are in extracting concessions from a counterpart. As a start, size does matter! Selecting the right size has two dimensions, the costs to the issuer and the cost/benefit to target.

Let’s begin with promises. The size of the promise is an important determinant for both the issuer and the recipient. From the perspective of the issuer, the lower the cost of living up to the promise and the higher the reputational costs of reneging, the more credible the promise.

Given that the issuer is weighing the cost of the reputational damage against the cost of fulfilling the promise, the best promises from the issuer’s perspective are those that are cheap to carry out, have high reputational costs of reneging, and have large benefits to the target. A good example is for a car dealer to promise superior service to customers who bought the car at the dealership (putting them ahead of those who bought their cars elsewhere or providing complimentary car detailing after service) and hiring an independent agency that will advertise the service experiences of its customers. The costs to the dealer are low relative to the reputational costs of reneging, and the benefits to the customers may be sufficiently high that they would be willing to pay more for the car in their negotiations.

A similar approach can be used to determine the optimal size of a threat. The issuer can make threats more credible by choosing threats that are cheap for the issuer to carry out, perhaps because they provide a large amount of schadenfreude, while inflicting relatively large costs on the target. For example, the threat to key your car (a small threat, typically a misdemeanor if caught) is more likely to be carried out than the threat to kill you (a really large threat that may result in life imprisonment). Thus from the perspective of both the issuer and the target, threatening to key the target’s car is more credible than threatening to kill him. In contrast, in the case of the incumbent firm that tries to dissuade potential entrants, issuing the threat and fighting a price war in a very public manner, while costly, may be effective at deterring future start-ups from entering their market. Thus, what may be viewed as a crazy decision in isolation may actually be a very rational course of action when considered in a broader context.

In addition to size, two additional factors have been investigated: the form of the threat and the timing. You can issue an explicit threat or an implicit threat. An example of an explicit threat is “If you don’t concede, I will inform your superior about your aggressiveness.” An implicit or vague threat does not suggest a specific action; rather it implies that something bad will happen. “If you don’t agree to my request, you will regret it.”

How effective implicit or explicit threats are depends on when they are issued. Explicit threats are much more effective if they are issued late in the negotiation. Vague threats are more effective if they are issued earlier. Conversely, issuing an explicit threat early in the negotiation or a vague threat late has a worse impact on your counterpart’s willingness to concede than their response would have been if you had issued no threat at all! Issuing the wrong type of threats at the wrong time is not only ineffective, but it also makes you appear much weaker or more aggressive than you’d probably like.6

Up until now, we have been discussing the credibility of promises and threats when no further interactions among the negotiating parties are likely. Changing the scenario by adding the possibility of future interaction does not change any of those conclusions, but it does make threats and promises more credible. The reason is simple: the presence of a future makes reputation a much more important consideration.

When there is a future, the situation is more complex. First, the very fact of issuing a threat may negatively affect your future interactions. Having a future favors promises when both promises and threats are an option. In addition, the possibility of future interactions increases both the likelihood of carrying out the threat if the target does not concede and the likelihood of making good on the promise if it does, because failing to follow through ruins your reputation in future interactions. Conversely, if you carry out the threat or honor the promise, you gain a reputational benefit of being tough (in the case of a threat) or trustworthy (in the case of a promise).

But the prospect of future interactions also has implications for the target. If the target yields to a threat, then the issuer gains a benefit in future interactions with the target because she knows about the target’s willingness to concede when threatened. In the case of threats, the costs to the target’s reputation are likely to reduce her willingness to yield because by not yielding, she establishes her own reputation for toughness. Of course, she is also likely to suffer the consequences of the threat and the issuer’s enhanced reputation for toughness as he implements the threat. Therefore, when issuing a threat when there is future interaction, you must expect an increased probability of the target not yielding, and hence should seriously consider the situation you will face: not only will you not get the hoped-for concession, but you also may have to carry out the threat—or suffer reputational costs—if you do not.

Consider how the challenge of a future influenced the unsuccessful strategy of one of our colleagues. He received an offer from a prestigious east coast business school and proceeded to go to the dean with a threat: either match the offer, or I will leave. Given that our faculty colleague had a competing offer, his threat appeared credible; the dean, the target of the threat, could have complied. However, doing so—especially when he knew that his behavior would become known to other faculty members—would set a costly precedent by indicating to the rest of the faculty that the dean would raise their salaries when threatened. So the dean simply responded by wishing our colleague well in his new job. Although the dean valued the individual contributions of our colleague, the cost of complying with the threat would have been not only the cost of the additional compensation to this faculty member, but also the compensation of all the other faculty members who would inevitably have lined up outside his office with their new offers in hand.

In the case of threats like this, the issuer must also consider the reputational costs of not following through with a threat. When our colleague stayed at Kellogg, despite the dean’s not matching his offer, no one, especially the dean, took his subsequent threats to leave seriously. Therefore, especially when future interactions are likely, the issuer should first consider whether to issue a threat in the first place because just issuing that threat may have negative consequences.

But as it turns out, there was a better strategy than issuing a threat. In negotiations with that same dean, a more skillful faculty member brought her competing offer to the dean and simply asked if the dean could let her know what her next year’s compensation would be. This approach differed in two important aspects from the first one. First, because it was not formulated as a threat but rather as a request for information, it did not provoke a natural concern that the dean would get a reputation of yielding to threats. More importantly, it reduced (even though it did not completely eliminate) the pressure on her to accept the outside offer or risk losing face. What if the dean were to counter with an offer for the next year’s compensation that was significantly lower than her competing offer and our faculty member were to remain at Kellogg? While the faculty member would have to make a choice, her behavior would not have painted her into so obvious a corner. But in this case, the dean matched the outside offer, and the faculty member happily remained at Kellogg. The whole negotiation took less than ten minutes.

THE POWER OF PRECOMMITMENTS

Consider the car dealer’s promise to provide you with superior service after the purchase service. In general, you might not take that promise seriously if it were issued privately. But the dealership can make this promise more credible not only by making it publicly but also by hiring an independent marketing firm to follow up with customers, polling and publishing their after-the-sale service experiences. By entering in such an arrangement—a precommitment—the dealership effectively increases its reputational costs of reneging and hence makes the promises more credible. In turn, customers rationally will view such promises as more credible and be willing to make larger concessions, for example to pay more for the cars.

A similar precommitment can be used to make threats more credible. One example occurred during World War II. Switzerland was surrounded by Germany and its allies. It realized that it had a strategic importance to Germany because its passes over the Alps provided efficient routes to transport men and equipment between Germany and Italy, Germany’s major ally. Of course, Germany could make that transport even more effective by occupying Switzerland. Although the Swiss Army was a moderate deterrent, realistically it could not have withstood a serious attack by the Wehrmacht.

To deter a German invasion, the Swiss mined all (both in the interior and on the border) their country’s passes, tunnels, and bridges with explosives. Yet, on its own, the threat to blow up those passes and bridges would not have been credible. Once Germany invaded, after all, what would be the point of blowing up all the passes, tunnels, and bridges in the country? To give their threat credibility, the Swiss command assigned each bridge, tunnel, or pass a small military unit that had an irrevocable order to blow it up at the first sign of a German invasion anywhere in the country. If the officer in charge refused to obey the order, the men of the unit had the order to shoot the officer and detonate the object. Then the Swiss commanders made it easier for the unit to follow its orders. They assigned to each object troops who were not local. That is, units from the German part of Switzerland were assigned to the bridges, tunnels, and passes in the Italian part; while those from the Italian part of Switzerland were assigned to guard the French part; and so on. In this way, the army took advantage of Switzerland’s cultural diversity to make it easier to follow the order because, in the sense, each unit would be blowing up somebody else’s property.

Of course, the “secret” plan was then leaked to Berlin. In effect, what the Swiss strategy did was to put the decision to detonate the passes, bridges, and tunnels in the hands of the Germans. If they invaded Switzerland, they would lose what they wanted most. If they did not invade, they could still transport men and equipment more efficiently through a Switzerland with bridges, tunnels and passes than through a Switzerland without them! What both of these stories have in common is that a precommitment removes the decision that the issuer must make to actually implement the threat. When the commitment is irrevocable, it is more credible.

SUMMARY

When you and your counterpart have a future together, issuing threats exposes you to two negative consequences. First, there is the negative consequence of issuing the threat. Second, not carrying out that threat will damage your reputation. In such situations, therefore, you should only issue threats that you are ready to carry out. But again, big threats (ones that impose large costs on the issuer) may still not be credible if the reputational costs are less than the costs of carrying them out. Thus, our recommendation still holds: issue the smallest threat that is large enough to get your counterpart to concede.

The same is true for promises. Although issuing promises per se does not negatively affect future relations, the prospect of future negotiations makes it less likely that the issuer will renege and thus makes promises more credible. Of course, the issuer might still renege if implementing the promise is costly. Thus, for promises to be credible, the cost of making good on them to the issuer must be less than the reputational cost of reneging. So, consider making the smallest possible promise that is enough to make your counterpart concede. These smaller promises and the prospect of future dealings will make the promises more credible and thus more influential.

What emerges is the following decision pattern:

• Without the expectation of future interactions, privately issued threats or promises are not credible, should be ignored, and hence should not be issued. However, threats may be used by the target to retaliate (schadenfreude!) by making them public (even if they have not been carried out). Therefore, threats are not only unlikely to be effective without any future interactions, but they can also impose significant costs on the issuer.

• Without the expectation of future interactions, the issuer can increase the credibility of threats or promises by issuing them publicly or committing to make their fulfillment public. This may be sufficient to make them credible when the reputational costs to the issuer of not following through are high and/or the costs to the issuer of following through are low. Therefore, threats that are minor relative to the reputational loss of carrying out the threat, issued by a party with a high reputation, should be taken seriously while large threats or promises issued by parties with a low reputation should be ignored.

• With the expectation of future interactions, threats or promises may be credible when the reputational costs for future interaction to the issuer of not following through are high and/or the costs to the issuer of following through are low. Therefore, little threats or promises issued by parties with a high reputational capital should be taken seriously, while large threats or promises relative to the reputational costs of reneging should be ignored. Again, the issuer can make both threats and promises more credible by making them publicly and also committing to make it public that they have been carried out. Finally, issuing threats per se—even credible ones—may negatively impact on future interactions.

• Before yielding to a credible threat or promise, you still have to assess whether the concession is more costly to you than suffering the cost of the threat or benefiting from the promise.

To assist your assessment of your actions, regardless of whether you are potentially the issuer or the target of a threat or promise, we have created Matrix 9.1, which summarizes how effective a particular action might be.

The conclusion is quite straightforward: threats and promises will not be effective unless the target expects them to be fulfilled. How likely they are to be fulfilled rests on the relative cost to the issuer compared to the cost of reneging. So clearly both issuer and target need to consider when a threat or promise is credible and when it is simply cheap talk!