Sources of Power in Public Negotiations: A Framework Applied to Public‐Public and Public‐Private Negotiations

Power is one of the most often‐discussed topics in any treatise on negotiation, and the frameworks provided for analyzing it play a central role in the preparatory phase of a negotiation. These frameworks provide a useful model to guide individuals as they catalog the various sources of power available to them and their counterparty in a negotiation. In addition, analysis of power dynamics offers a roadmap for an individual to use in attempting to mitigate power imbalances. However, a leader in the public sector might find the current models lacking when it comes to understanding the nuance involved in a public‐public or public‐private negotiation. Most current frameworks for power were constructed with a sector‐agnostic or private‐sector focus, and while they can be stretched to accommodate the unique circumstances faced by governmental actors, it isn’t a natural fit. As a result, we believe a re‐configuration of current frameworks for power in negotiations can offer a clearer insight into the power dynamics at play for civic leaders or for individuals negotiating with governments.

One particularly salient example in recent memory is the breakdown in negotiations between Amazon and New York City for the establishment of a second headquarters. After spending a year searching for a new headquarters, Amazon ultimately decided in November 2018 to split their expansion between two different locations: Crystal City in Arlington, Virginia, and Long Island City in New York City (Weise and Goodman 2018). Almost immediately after the announcement, however, problems began to develop with respect to the New York City campus. Local residents began to raise concerns about the new strain on transit and sewage systems caused by tens of thousands of new workers on the Amazon campus, and the local community board worried about businesses being pushed out by rising land costs (Goodman 2018). However, Amazon pressed ahead, and judging from its actions believed that its early support from key politicians such as New York’s governor, Andrew Cuomo, and the city’s mayor, Bill de Blasio, would prevent any issues from arising (Roose 2019). However, this attitude ignored key power brokers involved in a deal of this scale in New York City, as well as seismic shifts underway in New York state politics. Furthermore, Governor Cuomo and Mayor de Blasio made many of the same mistakes as Amazon did, apparently trying to make an end run around the City Council and normal land‐use procedures in their eagerness to get a deal done (New York Times Editorial Board 2018).

As a result, Amazon found itself pulling out from its planned expansion to Long Island City only three months after the announcement—despite widespread support for the expansion among New Yorkers, the mayor, and the governor. However, looking solely at these obvious indicators ignores several other power brokers whose influence ultimately led to Amazon’s decision to leave (Goodman 2019).

Just the weekend before Amazon’s announcement, the company’s executives found themselves in a fight with a suburban state senator with the power to block Amazon’s plans through the use of an obscure state board, the Public Authorities Control Board. The neglect of Amazon, alongside Governor Cuomo, to examine the board’s veto power resulted in a failure to engage with key constituencies in the weeks and months prior to the announcement. Cuomo appears to have made a key mistake in realizing the importance of the board only after an Amazon opponent had been appointed to it. At that point, he was left with the unenviable option of letting the process play itself out or utilizing his own precious veto power to block the appointment (Goodman and Weise 2019).

Mayor de Blasio also appears to have underestimated his opponents’ power. He did not recruit Corey Johnson, the speaker of the city council, as an ally early on in the process. Johnson went on to damage the deal by holding numerous public hearings instead of voicing his concerns privately to Amazon. It was only in the final days of the deal that de Blasio began to strategize about bringing powerful local unions on board. Likewise, only belatedly did Amazon begin to rally support from community leaders, and the company never sent its head of press and government relations to New York to meet with officials, despite the mounting pressure. In examining the failure of the company’s strategy, Joseph Parilla, a fellow at the Brookings Institution, stated that “Amazon underestimated the power of a vocal minority and miscalculated how much it needed to engage with those audiences to make HQ2 a success.” Due to Mayor de Blasio’s and Amazon’s failure to bring Corey Johnson and the City Council on board early, Amazon endured hours of public excoriation during City Council meetings, which brought unwanted headlines and discussion about the company’s practices. During the meetings, Amazon officials seemed taken aback when protestors heckled them and when council members forced them to publicly state their anti‐union stance. Perhaps the most influential up‐and‐coming politician in the area, Alexandria Ocasio‐Cortez, was a vocal skeptic and her attitude spread to her supporters. Although these issues did not implicate any kind of formal or institutional power that could prevent Amazon from breaking ground on its headquarters, they did create public relations challenges at a time when the public was becoming increasingly skeptical of “big tech” (Goodman and Weise 2019).

It is difficult to put the sources of power that proved so troublesome to Amazon and its New York allies into the traditional categories used by social scientists and the negotiation academy. Would a veto by Governor Cuomo be considered coercive power, or legitimate power? Likewise, would the protestors’ and combative city councilors’ ability to create a nuisance be categorized as coercive power, or contextual power? These two examples are embedded easily within a model of power crafted for this sort of negotiation, however. The Public Authorities Control Board, and the state senate, by extension, wielded veto power in their negotiations with Amazon, and the city council members and their supporters possessed easily deployable nuisance power. These are just two examples of the types of power we identify as being particularly relevant to negotiations in the public sector.

These negotiations do not just prove perilous for private actors seeking to negotiate with the government, however. Misunderstanding of the source of power at play can also affect governments when they attempt to enter public‐private partnerships, to the detriment of the public. California experienced just such a suboptimal outcome with a public‐private partnership to create a four‐lane toll road in 1995. After initial success, some of the terms in the contract, such as a thirty‐five‐year noncompete clause preventing the state from repairing the road or adding lanes, proved untenable. The resulting litigation took a decade before it was resolved, and in the end the state was forced to buy back the lanes from the company in order to make needed changes (Goldstein and Cohen 2017). Given the current infrastructure crisis in the United States, it’s likely that such public‐private partnerships will proliferate, especially given their warm reception by the most recent two presidential administrations (Ivory, Protess, and Palmer 2016). As such, it is important that public officials representing cities and states prepare effectively in order to create mutually beneficial agreements with private‐sector counterparties.

In this article, we will explore the importance of power in negotiations and the ways in which it is traditionally defined, before presenting a new model that can achieve the same objectives while better grasping the nuances of public negotiations. From there, we will examine a recent case study of a public negotiation through the lens of these new sources of power, in order to show how the model might be used by the parties in such a deal.

The concept of power has been a central part of the negotiation canon since its inception. The “most often quoted scholars on power” predate the field itself, as Bertram Raven and John French wrote about the topic in 1959, before the emergence of the academic study of negotiation in earnest (Lewicki et al. 2014: 257). Power has earned such an important place within the academy because it is seen as giving a significant advantage to the party who possesses it. If a negotiator feels that they have less power than their counterparty, they are likely to seek power to offset their disadvantageous position or to counterbalance the other party in some other way. Likewise, a negotiator with more power than their counterparty may believe that this advantage will allow them to secure their desired outcome, and will attempt to sustain their advantage during the negotiations that follow (Lewicki et al. 2014). As a result of this, the concept of parity in power has taken on great importance. If two sides believe their power is relatively equivalent, they will know that the counterparty will be able to counter their power, rendering escalation useless (Pienaar and Spoelstra 1992).

However, power itself is a somewhat tricky concept to define. French and Raven state that the concept of power can only be useful “if it has a certain stability over time; it is useless if every momentary social stimulus is viewed as actualizing social power.” (French and Raven 1959: 152). They examine the topic from two separate vantage points: first, how power determines the behavior of an agent exerting it; and second, how power determines the reactions of the recipient of the behavior. Further, they separate out power from influence—which they see as power operationalized or put into use (French and Raven 1959) (Table One).

French and Raven identified five key “bases” of power; these five sources remain the foundation for many modern models: (1) “reward power;” (2) “coercive power;” (3) “legitimate power;” (4) “referent power;”and (5) “expert power.” It is useful to define each of these types of power in turn, as their echoes can be seen in their more modern relatives. Reward power is perhaps the most self‐explanatory of the group, as it is defined simply as “power whose basis is the ability to reward,” with the strength of the power increasing with the magnitude of the reward a counterparty perceives. Perhaps reflecting the prevailing focus of organizational research of their time, French and Raven offered the example of a bonus given by a manager in a factory as a basis for reward power. Coercive power is similar in many ways, but differs primarily due to its basis in the ability of one party to punish another. The power stems from one party’s expectation that they will be punished if they “fail[] to conform” to the other party’s attempt to influence them. The preceding two powers can naturally blend together, which can make disentangling them difficult—as French and Raven note, there is some difficulty in distinguishing whether the “withholding of a reward” is actually “equivalent to a punishment” (French and Raven 1959: 150–157). This is a dilemma that future models of power would attempt to resolve.

Following the relatively straightforward first two sources of power is the slightly more nuanced concept of legitimate power. This source of power is derived from internalized values that lead one to believe another has a legitimate right to influence them. In fact, this source of power has roots in scholarship on the legitimacy of authority preceding French and Raven, but they expanded it beyond those earlier concepts. For instance, there need not always be a role relation (of superior and subordinate, for example); legitimate power could also follow from a previously made promise by one party. An individual’s adherence to a code or standard (such as honoring one’s obligations) could therefore allow a counterparty to assert power (and enforce the previously made promise). At the core of legitimate power is belief in a system—whether it be a moral system, a culture, or a hierarchical structure in an organization. For example, an elected official only has power to the extent that her constituents accept the outcome of the election as legitimate (French and Raven 1959). Legitimate power comes perhaps the closest to explaining some of the nuances of negotiations in the public sector, but the sweeping nature of the term makes its practical use somewhat limited.

Perhaps even harder to define than legitimate power is referent power, which is based on the identification of a person with a group, an individual, or an idea. The nature of that relationship gives some power over the person, although it can easily drift into territory already covered by reward and coercive power. Conforming to the norms of a group only because you fear ridicule would be based on coercive power, not referent power. And similarly, conforming only to obtain praise would be based on reward power. Therefore, it can be difficult to find discrete instances of referent power; this is perhaps in part the reason the concept was updated by later scholars. French and Raven note that referent power can be created through the utilization of actual rewards over time, creating a stronger identification between two individuals that becomes independent of the reward (French and Raven 1959).

The last of French and Raven’s categories is expert power, alternatively called information power by contemporaneous researchers. This power is created when a person perceives another to be an expert within a given area. Further, expert power is relational, as any individual will evaluate another’s expertness against his own expertise as well as against an absolute standard (French and Raven 1959). A classic example of expert power would be an independent party brought in to a negotiation to appraise an item—the latent expert power would be utilized to influence one party or the other to change their stance on the price.

Since French and Raven’s initial scholarship on the topic of power, there have been numerous other attempts to categorize power as it relates to negotiations. Some scholars have focused on the best alternative to a negotiated agreement (BATNA) as the single “most important source of power in a negotiation” (Thompson 2012: 154). The availability of a BATNA allows a negotiator to make decisions not on an absolute basis, but as compared to an alternative deal that is available (Lewicki et al. 2014). Under this model, the most important assessment a negotiator should make is how their BATNA compares to that of their opponent, and how they could improve the differential (Thompson 2012). Other members of the negotiation academy have attempted to update and revise the bases of power model to apply to a wide range of negotiations. One particularly comprehensive model, which we will now examine, is put forth in the treatise on negotiations compiled by Roy Lewicki, David Saunders, and Bruce Barry (Table Two).

Similar to French and Raven, power is split into five distinct categories in this updated model, but the categories themselves are quite different from the original bases of power. The new categories proposed by Lewicki and his colleagues are: (1) “informational power;” (2) “power based on personality and individual differences;” (3) “position‐based power;” (4) “relationship‐based power;” and (5) “contextual power” (Lewicki et al. 2014: 263).

Informational power can clearly be compared to expert power from the old formulation, but with some major differences. Rather than being based on the expertise of an individual, informational power is based on the ability of a negotiator “to assemble and organize facts and data” in support of her position. This is highlighted as the “most important source of power” in this new framework, given its ability both to bolster a negotiator’s position and undermine the other side’s negotiating argument (Lewicki et al. 2014: 264). Power flowing from the expertise of an individual is seen as a subset of information power, functioning in much the same way as in French and Raven’s original model.

The next category of power, power based on personality and individual differences, dives deeply into psychological factors that might serve as a source of power for individuals in a negotiation (Deutsch 1985; Lewicki et al. 2014). The focus for this category is on “stable individual differences” that determine how an individual might behave in any given situation. They include an individual’s orientation toward power generally, their motivations, and their disposition (Lewicki et al. 2014: 266–267). An example of this in action could be seen in an individual with a strong moral commitment to the philosophy of egalitarianism entering a negotiation. All other things being equal, such individuals have been shown to be more likely to share power with low‐power parties (Coleman 1997).

Position‐based power has some similarities with the concept of legitimate power, but it expands the idea to encompass a broader set of organizations than the older idea. In addition to the idea of legitimate power that flows from a specific title in a hierarchy, position‐based power includes “resource power” that is based on the “control over resources” that go along with a certain role. In addition to making several of the same points about legitimate power as French and Raven, Lewicki and his colleagues note that individuals who hold legitimate power often amass resource power as a way of “fortify[ing] their power base.” This shift perhaps acknowledges the shift of power within organizations from being based purely on hierarchy to being based more on the value of resources that one can bring to bear. Embedded within resource power are the older concepts of reward and coercive power—an individual is able to deploy or withhold resources in order to influence others. One of the most well‐known examples of a figure with a significant amount of resource power but very little legitimate power is Robert Caro, the famous power broker of New York City who used control over tolls and other hidden resources to amass enough power to reshape the landscape of the entire city while never actually holding elected office. Related to this concept of resource power is the idea of network‐based power, a further subset of position‐based power. This type of power is a response to the move of organizations away from a strictly hierarchical setup into a “network of interrelationships.” Under such a model, it would be difficult, if not impossible, to determine the power of an individual based on their title. Rather, the correct way to examine such power would be to look at the individual’s ties to others within the organization, the strength of those ties, and the nature of the relationships. Some relationships might be based more on personal connections, whereas others might be based on, for example, an exchange of resources (Lewicki et al. 2014: 268–275).

The next category, relationship‐based power, has some overlaps with the network concept above, but is more similar to the older category of referent power. A relevant example of referent power to the public setting would be the negative referent power that exists when political rivals negotiate “across the aisle.” Even when interests are aligned, the ideological differences and past political disagreements can create forces that work against an agreement. Beyond just the referent power gained from a close association, though, this concept of relationship‐based power is expanded to also include power based on goal interdependence. Two parties with goals that can be achieved in cooperation with each other would be less likely to utilize power against each other, as compared to parties in direct competition with each other (Lewicki et al. 2014).

Finally, this new formulation ends with the concept of contextual power. This type of power is perhaps the most novel in the new formulation, and also encompasses the BATNA‐focused approach to analyzing power. Contextual power can be created by the “context, situation, or environment in which negotiations take place” and as a result can serve as a catch‐all for many other types of power that might impact negotiations. An elected official’s constituency could serve as a part of the overall context for a negotiation, just as the presence of protestors at city council meetings (Lewicki et al. 2014: 278–279). Therefore, numerous additional sources of power relevant to public negotiations could be included as part of this source of power.

Although the reformulation of power proposed by Lewicki and his colleagues provides a more useful framework than the original conceptualization of French and Raven, there are still numerous challenges in trying to fit the unique circumstances of a public‐sector negotiation into its categories. In the next section, we will examine some of those challenges, and propose a new formulation of power specific to this type of negotiation.

Scholarship on negotiation and strategy frequently includes examples drawn from the world of government and diplomacy. The early chapters of several treatises make reference to just such situations, referring to countries negotiating “to open their borders to free trade” (Lewicki et al. 2014: 3) or the “allocation of land” in the Israeli–Palestinian conflict (Thompson 2012: 29). However, while these texts make reference to such circumstances, many of the practical exercises are drawn from everyday situations in the workplace, such as negotiating a salary or contractual terms. As a result, there is an emphasis on these topics, which helps explain the difficulty of using traditional power frameworks for public negotiations.

Although much negotiation literature examines workplace or organizational issues, there has also been research examining negotiation through the lens of diplomacy and grand strategy and such works can offer insights into how power might be defined in such public negotiations. Thomas Schelling’s classic work The Strategy of Conflict offers several insights that could be integrated into a more specific definition of power. Schelling looked beyond traditional definitions of power, focusing instead on a particular type of tactic: the “irreversible sacrifice of freedom of choice.” In making such a decision to “bind oneself” a weaker party could often use an opponent’s strength against them, such as in the case of union negotiators who enter a negotiation with the union already having voted to authorize a strike if their demands are not met. Beyond this general idea of power, however, Schelling also discusses some of the structural aspects of a negotiation that might affect the power of each party. Use of agents with limited authority can affect each party’s ability to change the terms to which they could acquiesce (much like the above example of “binding oneself”). Additionally, the format of the negotiation itself can have a significant impact. If a negotiation is conducted in public, the party that has a “public” to whom to answer will be significantly constrained as compared to a negotiation carried on in secrecy. Further, even if the proceedings are secret, the party without a relevant “public” will have the ability to publicize or leak information, providing an additional source of power (Schelling 1960). The power of the “public” is quite evident in Amazon’s failed plan to locate a new headquarters in Long Island City, as negotiations proceeded quite smoothly while they occurred in secret. Only after the details of the plan became public did previously neutral or friendly legislators begin to question the wisdom of the plan.

Schelling also discussed the idea of intersecting negotiations—that is, if one party to a negotiation is engaged in simultaneous negotiations, it can persuasively argue that its reputation will be affected by any actions that it takes in its current negotiation. This is as compared to a party engaged in negotiations with only one counterparty, where such an argument would have little persuasive value. Continuous negotiations between two parties have much the same effect—one negotiator can stand firm by arguing that a concession would affect the parties’ future negotiations. Such an argument bolsters the credibility of the party and confers additional power that might previously not have existed. An additional tactic that can provide power to one party or the other is the ability to restrict or expand the agenda for a negotiation. By limiting the scope of topics that a negotiator has authority to decide, a party can be protected from extortionate measures by an opponent (Schelling 1960). The power of the agenda can be seen in the failure of Amazon’s deal. Had Amazon been able to cabin the topic of tax credits and focus solely on solvable issues of infrastructure and land use, they would have been in a much more advantageous position.

Another important insight discussed by Schelling is the power granted by the “mechanics of negotiation.” This implicates the importance of the veto power in the Amazon case, which we discussed in the introduction. Some of the examples of such mechanics that are offered by Schelling are parliamentary techniques—such as the veto—and the structure of auctions—for example, a Dutch auction as opposed to a sealed bid system. These mechanical features can certainly alter the balance of power, based on the specific feature and the person who is able to decide whether or not to utilize it (Schelling 1960). Given the importance of such mechanical constructs to governmental decision‐making processes, it’s quite evident why they would be so important in negotiations in the public sphere.

Finally, Schelling discussed the power conferred by threats in negotiations—threat communication. The threats that he discussed are generally focused on deterrence, given that both parties would generally be better off if the threat itself was not actualized. Rather, the important thing, according to Schelling, is for a party to demonstrate that it has bound itself to “fulfill the threat,” thus increasing its leverage. Schelling asserted that a more effective threat is one that can be “decomposed into a series of consecutive smaller threats” to allow a demonstration without instantly terminating a deal (Schelling 1960: 35–41).

These concepts from the domain of strategy provide several building blocks that will factor into our new framework of power for public negotiations. However, some additional writings from the negotiation academy can help shine further light on how power might be better catalogued for such negotiations.

In The Expert Negotiator, Raymond Saner takes an approach to negotiating power that is different than those of the more traditional frameworks presented earlier. He splits negotiating power into two categories—the power that comes from “the situation itself” and the power that participants possess as individuals. Situational power is largely fixed before the negotiation begins and includes each party’s BATNA and access to information. By contrast, personal power is derived from the traits that each individual brings to the negotiation, such as expertise in a field and a personal ability to levy rewards or punishments (Saner 2012). This emphasis on the importance of the situation is apt given that Saner’s treatise focuses on international affairs, as opposed to the more commercial nature of other texts.

Jeswald Salacuse has written about specific “power tools” that may be brought to bear in negotiations with governments, identifying such tools from real‐world negotiations and framing them within broader negotiation theory. The first tool is the “power of precedent,” wherein a past example may be “used to justify or guide” a government’s action in a similar negotiation. His example of this type of power involved a government grant to a private organization of an exemption to a statute or regulation. The past precedent was a significant source of power for the individual in the negotiation, as it provided guidance arguing in favor of the exemption. Salacuse attributed this power to the need for government negotiators to “demonstrate that they are treating people fairly” and the career‐threatening implications of taking actions that may give the contrary impression. Related to the concept of precedent is the desire to avoid “surprises.” This was demonstrated, Salacuse explained, when the “approval by town regulators of a project that unexpectedly sets off furious community opposition is a surprise” (2008: 133). The parallels to the Amazon situation are apparent—presumably Amazon’s allies in the government, including Mayor De Blasio, were not pleased with the unexpectedly loud and angry response from their constitutents. Although such surprise tactics might seem like a potent source of power for an individual, Salacuse cautioned against using surprise tactics against a government counterparty given the potential of poisoning relations with them in the process (Salacuse 2008). Implicit in this analysis is the importance of understanding the second‐order effects of any deal and the reactions by relevant constituencies. Had Amazon and its allies within New York better understood the local environment before announcing the HQ2 deal, they might have avoided the resulting “surprise.”

Saner’s and Salacuse’s discussions of power offer several new ways of thinking about the topic that will influence our new framework on power. Perhaps the most important idea is that power operates differently during negotiations in which the government is involved and should therefore be examined differently. The framework that follows will attempt to provide a new lens through which to view such negotiations.

In reviewing the literature on negotiations involving governments, a few themes emerge. One prominent theme is the importance of formal power structures given the institutional bend toward them in the government. The ability to limit the range of possible outcomes through the veto power or the restricted power of agents, as highlighted by Schelling, is often at play in governmental negotiations. However, informal power also plays a key role. As seen in the Amazon negotiation and other situations discussed by Salacuse, bad publicity or other unexpected events can drastically shift the power balance in a negotiation. Finally, some power incorporates elements of both formal and informal mechanisms, and can be seen as drawing on a mix of the two (Table Three).

Based on the previous insights, we propose a new model for categorizing power in public negotiations based on formal, mixed, and informal sources of power. Many of the types of power we include are adopted from previous research by Lewicki and his colleagues, categorized in a new way. In addition, in deciding to use this new breakdown, we draw partly on Saner’s insight that a different categorization can help provide insight into a different type of negotiation. Further, this framework helps to fill in some of the gaps in explaining the dynamics of a negotiation involving a governmental entity with significant institutional constraints. As such, we will begin our explanation of this new model with the category of formal power.

One of the themes running through Schelling’s writing on strategic bargaining is the importance of formal mechanisms, which we have grouped into four subcategories. Institutional power refers to the formal power that an individual is granted by virtue of their position. In many ways, this type of power is also similar to the concept of legitimate power as advanced by French and Raven, but the term is not as all‐encompassing as their use. Rather, institutional power refers only to the formal authority that an individual possesses based on their position within a governmental structure. The power that is related to that, such as the power granted by a certain culture or deference to authority, is not contained within our narrower term.

In addition to institutional power, formal power includes the concept of shift power, which is also influenced by Schelling’s work. Schelling noted that with a pre‐bargaining agreement, negotiators for a union could effectively tie their hands so as to limit the available alternatives they could accept. Likewise, an organization could provide agents with only limited authority for a negotiation in order to give them an edge in bargaining. However, shift power is the inverse of the sources of power discussed by Schelling. Shift power refers to the ability to switch between multiple types of acceptable deals without incurring a high personal cost. Thus, someone who has limited room to maneuver based on preexisting agreements or statements would possess low shift power. While Schelling pointed out that this can be advantageous, particularly in single‐issue negotiations, in the complex world of public negotiations it can often limit a public official’s ability to be flexible across multiple issues and parties. However, low shift power can often translate into more moral suasion power, which we shall discuss later.

The next formal source of power is convening power, which relates to the ability of a party to bring groups together based on an agenda over which the convening party has control. As before, this ties to the lessons of Schelling, in particular as relates to the importance of controlling the agenda for a meeting. By limiting or expanding the list of discussion topics, a negotiator can gain power over their counterparty. This is particularly salient in governmental negotiations, given the importance of parliamentary procedures and public hearings prior to deal approval. Amazon’s experience in New York city council meetings shows just how pivotal such events can be, and how much power is wielded by the party in the driver’s seat. Despite Amazon ally Mayor de Blasio’s power within the city, it was ultimately the Speaker of the City Council, Corey Johnson, who held convening power and used it to hold damaging public hearings.

The final source of formal power that we have identified is veto power, which refers to a party’s ability to stop a deal unilaterally, often through a legislative process. This can be tied back to the idea of threats as advanced by Schelling, but the more important influence weighing in favor of this category is the large number of examples where such a veto has been deployed. The Amazon case is but one example of the threat of a veto dramatically changing the negotiating field; the case study set forth in this article will further elaborate on the importance of such power. The ability of a single party to completely derail a deal is an important part of negotiating in the public sphere, and as such should be a part of any power analysis.

Between formal power and informal power lies a category that blends the two together, which we shall refer to as mixed power. The first subcategory we shall examine is momentum power, which is defined as the power gained when one’s alternative is perceived to be making progress over the other options. This may come about because of formal steps that have been taken—such as the earmarking of funds for a project—but also is influenced by the informal support of different constituencies based on perception. This can be tied to the discussion by Salacuse on the importance of precedent and predictability in governmental negotiations. Given an option that appears to be dominant and has already had some vetting, many in the public sector are likely to rally behind it.

Coalitional power also serves as an important blend of the formal and informal categories, as it depends on a mix of ability based on formal authority and personal relationships. Typically, this power stems from a party’s ability to tap into a diverse group of allies and direct their resources to a larger cause. Such power may be based in part by a position in a hierarchy or a network, but also is based on personal, informal outreach, and as such blends the two categories. The opponents of Amazon could be said to have high coalitional power given their ability to recruit allies with formal power alongside informal groups that could agitate for change. Looking at the coalitions a party can bring to the table is an important step in any power analysis.

Finally, informal power also serves as an important part of the negotiation landscape for public actors. Though less clearly reliant on the structures of government than the sources already described, it is nonetheless unique to negotiations occurring in the public sphere. Moral suasion power, which is based on the ability to appeal to moral, ethical, and community values, is the first type of informal power that we shall examine. This power stems from the type of debate that often occurs in a public negotiation, given the nature of the deals that are being made. Negotiations that implicate controversial topics can often turn on moral or ethical arguments, as opposed to pure economic considerations. Further, this power is available to weaker parties that might otherwise not have a seat at the table. This is exemplified by the Amazon case, as community groups with no formal mechanisms to alter the deal nonetheless possessed a great deal of moral suasion power.

Nuisance power, which stems from a party’s ability to draw attention to themselves even without a formal role in a negotiation, is quite similar to moral suasion power. This technique can allow a group that is otherwise lacking in power to call attention to a specific issue and potentially change the narrative. Further, the ability of a negotiating party to call forth other actors to wield nuisance power can be used as a potent threat, as described by Schelling above. Salacuse noted the attitude of governmental negotiators toward surprises and protests, critical responses by the media, or other negative publicity can have an outsize impact on negotiations. As such, a negotiator would be wise not to discount the power of players outside the formal power structure when preparing to negotiate in the public sector.

Two‐level games can also be important in the analysis of power. To use Robert Putnam’s example, national leaders have to concern themselves with domestic groups and coalition building (the first level) as well as foreign developments and domestic reactions (the second level). The complexity of such a situation is presented by the fact that “moves that are rational for a player at one board… may be impolitic for that player at the other board” (Putnam 1988: 433–434). We will see both levels of the game implicated in the case study that follows.

Negotiations between different governmental entities prove fertile ground for examining sources of power in negotiations. In the case of Manchester, New Hampshire, the negotiations were complex and occurred at both the state and local levels as the city attempted to respond to the opioid crisis in 2018 and 2019 (Roberts et al. 2019).

Joyce Craig was elected mayor of Manchester in 2017, becoming the first woman to serve in that role in the city’s 173‐year history. One of the consistent themes of her campaign was the need for a strong response to the crisis of opioid overdoses and chronic homelessness in the city. At the time of the campaign, the city was seeing an opioid death almost every three days, a rate that continued to rise after Craig’s election. In addition, due to the limited resources for drug addiction and homelessness outside of major cities in New Hampshire, Manchester found itself attracting many individuals with needs for long‐term drug treatment and transitional housing. All of this combined to create a true public policy crisis by the time Mayor Craig entered office in early 2018 (Roberts et al. 2019).

One of the important programs being utilized in Manchester and across the state to address the opioid crisis at the time of Craig’s election was the “Safe Stations” program. This program, spearheaded by Craig’s predecessor, turned every fire station in the city into a 24/7 resource center for individuals suffering from opioid addiction. However, the program was of limited efficacy despite the efforts of those involved, due in part to the extremely low level of state funding for drug treatment. As a result, wait times for the programs associated with Safe Stations were long, and resources were limited. In addition, Craig faced a new crisis immediately before she took office when she received word that Manchester’s largest treatment center for addiction had closed after financial mismanagement and an unsustainably high level of demand for its services. To make things more complicated, the treatment center, Serenity Place, was the only provider currently serving individuals utilizing Safe Stations (Roberts et al. 2019).

One potential source for new funding and support for Mayor Craig’s attempt to shore up Manchester’s efforts to combat the opioid crisis was the state government in New Hampshire. The state had created a new treatment model funded with a forty‐five‐million‐dollar federal grant, but efforts to increase state funding beyond the grant had faltered after Governor Chris Sununu vetoed a budget containing such expenditure. Craig had no additional levers to pull in this second level, and as such was forced to look mostly to local partners to shore up the Safe Stations program. Moving back to the local level, she engaged with Manchester‐based nonprofit organizations, other local service providers, and private donors to quickly create a “re‐boot” of the Safe Stations program, newly designated as Safe Stations 2.0. The new iteration of the program was able to utilize Lyft as a partner for transportation, along with new service providers based on the types of services each individual was seeking. However, even after saving the Safe Stations program, Mayor Craig still faced a significant challenge in formulating a response to the city’s rising population of homeless individuals (Roberts et al. 2019).

Although homelessness had long been an issue in Manchester, Mayor Craig found herself in conflict with the city’s Board of Aldermen—Manchester’s legislative body—when it came to formulating a response. Craig convened a task force on the topic shortly after taking office, which included members of the business community and was chaired by a prominent nonprofit leader. Her objective was to improve service provision within the city and provide stable housing, while also addressing complaints about panhandling from residents and the business community. However, the Board of Aldermen worried about the resource drain that such a response would entail, given the large number of individuals from outside the city that came to Manchester seeking social services (Roberts et al. 2019).

The disagreement came to a head in March 2019 during a meeting of the board to discuss the homelessness crisis. One of the aldermen introduced an ordinance to allow business owners to remove sleeping or panhandling individuals near their stores, which passed by a unanimous vote of the twelve aldermen present. Mayor Craig ultimately vetoed the ordinance, as she worried that such a measure would go against her systematic approach and potentially lead to litigation with entities like the ACLU. Craig’s veto outraged the board, but she managed to persuade enough of the aldermen of the wisdom of her approach to avoid a politically ruinous override of her veto (Roberts et al. 2019).

Mayor Craig persisted in her efforts after her showdown with the Board of Aldermen, as she continued to engage with state and local officials to address the ongoing opioid and homelessness crises in her city. Her actions up to this point provide an excellent case study in power dynamics in public negotiations, which we will now examine.

The various constituencies involved in Mayor Craig’s efforts to combat opioid addiction and homelessness show the extent to which a public negotiation can differ from a commercial negotiation between private entities. In addition to working with the Board of Aldermen, Craig also had to consider the responses of the public, business owners, nonprofit organizations, and the state government. Each of these different actors impacted the negotiations differently, based on the power they brought to the situation.

Mayor Craig possessed a substantial amount of formal power after assuming office, although other actors also had access to such power. Simply by virtue of her position as mayor, Craig brought a significant amount of institutional power into each of the negotiations in which she participated. Her office allowed her to unilaterally influence the public agenda in Manchester—sometimes known as “problem nomination”—simply by elevating concerns. She was also able to use the power of her office to strategically broadcast or limit conversations, as needed. Craig was able to appeal to the public in the days leading up to the Board of Aldermen’s vote, building support for her position but simultaneously limiting the conversation to a smaller group of service providers when rebooting Safe Stations to ensure she was able to restart the program quickly. Conversely, Craig did not possess a significant amount of shift power given the importance of the two crises to her term as mayor. Because she had bundled the issues of homelessness and opioid use together and also had spoken about the moral and ethical importance of the issues, she had less flexibility than other actors to shift between multiple alternative strategies. By way of comparison, the Board of Aldermen and the governor had more shift power given their flexible stances on the issue, which in the case of the board allowed it both to vote in favor of an ordinance and then back down in the face of a veto. This response brings to mind a point made by Schelling—sometimes a leader can actually gain power by limiting their options (as in the case of Mayor Craig) although this necessarily entails limiting one’s access to shift power.

In addition to institutional power, convening power was an important source of power for Mayor Craig, as she was able to establish a formal task force on the issue of homelessness. This allowed Craig to control the task force’s agenda as well as its membership, and ultimately provided her with a strong ally in her clash with the aldermen. It is important to note that this differs from the less formal power provided by a coalition, although the two certainly work together. Finally, the importance of veto power was evident at both the local and state levels. Mayor Craig was able to use her veto threat to prevent a potentially problematic outcome while simultaneously gaining time to persuade the aldermen of the wisdom of her alternative. Likewise, Governor Sununu was able to use the veto as a way of forcing his legislature to go back to the drawing board to craft a budget. In both situations, the formal power of the veto was an important part of the negotiating process.

Mixed power was also on display during the negotiations in Manchester. Momentum power manifested itself in Craig’s response to the homelessness crisis. Her ability to form a task force and immediately respond to the problem created a large amount of momentum behind her favored alternative, weakening the position of the aldermen when they belatedly tried to derail it. Had Craig not come to the table with her own alternative already partially implemented, she would have had less power in negotiating with the neutral aldermen. Craig also wielded coalitional power when entering the negotiations, as she had assembled a diverse base of support for her proposals. As such, she was able to lean on those allies to reframe the debate and force her opponents to push back not just against Craig alone, but rather a diverse coalition including business owners, nonprofit leaders, and municipal officials. Both types of mixed power—momentum power and coalitional power—blended Craig’s formal authority with her personal relationships in the community.

The final category of power we will examine is, in many cases, the mirror of the first. Where formal power allows an official to use the tools of their office to gain an advantage in negotiation, informal power provides an alternative when those tools don’t prove effective. In Manchester, Mayor Craig’s moral suasion power was evident in the morally and ethically infused style in which she addressed the issues. This limited her formal shift power but provided another source of power in the moral force she brought to negotiations. This was especially effective given the context—the opioid crisis was broadly affecting communities not only in Manchester but across New Hampshire, and thus a rehabilitative approach was viewed more sympathetically. Mayor Craig might have lost this moral suasion power had she advocated for an alternative discordant with public opinion. Although it was not deployed, Mayor Craig also appears to have entered the negotiations with significant potential nuisance power. Had the aldermen ultimately decided to override her veto, she had several levers with which to put pressure on them based on the broad coalition she had assembled. Her supporters could have picketed, protested Board of Aldermen meetings, or taken other actions outside the formal power structure to influence her opponents.

These examples show the importance of examining power in a public negotiation through a specific lens. More general frameworks of power are certainly broad enough to encompass the examples listed above, but for a public official, it is less likely that they would provide a roadmap for exhaustively examining the power dynamics ahead of a negotiation. An evaluation of the formal, mixed, and informal power wielded by each actor is a helpful tool for a public official assessing the balance of power prior to a negotiation.

In this article, we have examined the role of power in negotiations and offered a new framework for negotiations involving the public sector. The existing frameworks for examining sources of power were critical to our undertaking and still serve as the primary resource for a negotiator thinking through the balance of power with a counterparty. However, we believe that this new framework helps to synthesize some ideas from the literature on strategy and negotiation between states to provide a better fit for public negotiations. The Amazon case provides a particularly salient example, given the high stakes involved and the clear importance of formal and informal power. It is impossible to know everything that the parties examined before announcing the deal, but future dealmakers would be wise to learn from this negotiation.

Power is clearly an important aspect of any negotiation. Negotiators whose power is matched to that of their opponents are far more likely to work toward a mutually acceptable outcome than those with a power imbalance (Lewicki et al. 2014). There are many different techniques for examining the power differential in a given situation but the main models that we examined in this article include French and Raven’s five bases of power and Lewicki et al.’s sources of power. These models provide insights into the power dynamics of a negotiation, and offer negotiators a way to analyze a situation before arriving at the bargaining table. However, these models do not incorporate all the existing research into strategic bargaining and interactions between different countries. In particular, Schelling’s research on strategy offers some insights into the power a party can gain through limiting its alternatives, utilizing parliamentary techniques, or threats. In addition, Salacuse and other authors have provided some examples of the specific considerations that factor into the decision‐making process for governmental actors, and how they differ from the concerns of their peers in the private sector.

Combining together these different insights resulted in our new framework for power, based on formal, mixed, and informal power. This model allows public sector negotiators to more accurately categorize and catalog the different sources of power at play, so they can enter a negotiation prepared. We examined the Manchester opioid case study through this lens, in order to show how it might be used by a practitioner in a real‐life scenario. The importance of sound negotiation practices in local government is more important now than ever before, given the increasing role of public‐private partnerships in municipal service provision. As a result, we hope that this framework will be of some use to public officials navigating the complex world of negotiation.

The authors would like to acknowledge the generous support of the Bloomberg Harvard City Leadership Initiative.