A review of suggested styles and studies of effectiveness of negotiation, and their application to commercial mediation
1. Aim of article
The genus of this article is finding myself in the position of trying to encourage participants in a mediation to go beyond stating their positions; encouraging them to explore their interests and, even more radical, to show curiosity about the interests of the other party. Commercial mediators’ training in the United Kingdom tends to suggest adopting the approach of Getting to Yes (Fisher and Ury, 1981), discussed below. But other descriptions of styles of negotiation exist that some participants, and indeed some mediators, may find more useful. And anyway, these are just academic suggestions; is there any evidence supporting the proposition that any negotiation style beyond ‘pitch it high and be awkward’ really works?
So my purpose is to explore leading academic work on the styles of, and behaviours in, negotiation, and to look at studies that investigate or measure their effectiveness. I then seek to extract conclusions that anybody trying to negotiate a settlement to a dispute or disagreement might use to inform their strategy and behaviour. I am particularly interested in whether a negotiator who shuns the more aggressive negotiation styles is likely to be worse off than one who adopts them. Finally, I am interested in how these conclusions can be applied in commercial mediation and whether the model of mediation generally in use in commercial disputes is well suited to their implementation.
2. Negotiating Styles
While the possibility of their being styles of negotiation may not have been invented in 1981, the publication of the first edition of Getting to Yes (Fisher and Ury, 1981) galvanised interest in the subject. The early 1980s produced a flurry of thoughtful work on which our current orthodoxy is still firmly based.
Before Getting to Yes
Prior to 1981 there was little in publications of guidance to negotiators that deviated from a traditional ‘hardball’ view of negotiating. As David Hoffman points out (Hoffman, 2003) Meltsner and Schrag (Meltsner and Schrag,1974) suggested a long list of manipulative and deceitful tactics that a negotiator should adopt, including
- appear irrational where it seems helpful,
- claim you do not have the authority to compromise,
- make your first demand very high,
- designate at least one demand a pre-condition,
- use good cop/bad cop,
- make the other side compromise first, etc.
Hoffman draws the conclusion that:
“The unspoken assumption in these suggestions is that the opponent is willing to take advantage of the negotiator — fairly or unfairly — and therefore success requires using competitive negotiation techniques, and using them more effectively than the opponent.” (Hoffman, 2003:p3).
But even before 1981 there were at least two areas in which something other than this pure competitive style was being considered. One of these areas was game theory, which had been a tool used to investigate negotiation tactics from the 1950s (e.g. Flood, 1952 and Dresher, 1961). This work involved looking at situations where a participant had to decide whether to stick to an agreement or renege on it, as in the Prisoner’s Dilemma. From this work Robert Axelrod (Axelrod, 1984) developed his theory of cooperation, further discussed below.
Another was the analysis of 1000 lawyers in the Phoenix area, done by Gerald Williams in 1976 and reported in Williams, 1983. In his survey he asked the participants to rate negotiators they had had to deal with in their last negotiation by reference to adjectives describing them (e.g. ‘astute’, ‘forceful’) and by reference to bipolar pairs of statements (e.g. ‘willing to stretch the rules’/‘not willing to stretch the rules’). From this analysis he derived two different styles, cooperative and competitive. Even more interestingly, Williams enquired after the perceived effectiveness of the negotiators, the results of which are referred to below.
Position-based – Principled
In Getting to Yes (Fisher and Ury, 1981) the authors argue for the supremacy of interests-based or principled negotiation over position-based negotiation. Position-based negotiation involves each side putting forward the position it is prepared to accept. Thereafter each side reiterates and/or moves its position without any explanation.
Fisher and Ury accept that position-based negotiation can partly meet one of their stated criteria for methods of negotiation, in that it can produce the terms of an acceptable (if not necessarily wise) agreement. But they contend that it meets this one only “eventually”, and so fails to meet one of the others: that it is efficient. It also fails to meet, they argue, the third, namely that it should improve, or at least not damage, the relationship between the parties.
Fisher and Ury subdivide position-based negotiation into soft and hard positional bargaining. The soft positional bargainer pays attention to relationships, sees negotiation as between friends and has a high priority on reaching agreement. The hard positional bargainer merely pursues what she wants, seeing the other party as an adversary and the aim as victory. In a contest between hard and soft positional bargainers, the hard one wins.
In contrast, Fisher and Ury put forward principled negotiation. This technique involves a number of key approaches that differentiate it from position-based negotiation. First is to separate the problem from the people who have the problem. The second is to focus on interests rather than positions. That is not to focus just on what each disputant is asking for, but on the reasons for asking for it and the needs that would be satisfied by having it. Another is to seek options for mutual gain; seek to expand the pie rather than just dividing it.
Distributive – Integrative
Bernie Mayer (Mayer, 2000), following others (e.g. Lax and Sebenius, 1986), prefers the terms distributive and integrative to describe aspects of negotiation, depending on their aim. Integrative negotiation is the style used when the negotiator is trying to meet his needs by making sure everyone’s needs are adequately addressed – that is those negotiating have the common interest of increasing the pie.
Distributive negotiation is about gaining as much as is possible of what is available. Unless, as Mayer puts it, “relatively unlimited amounts of benefits are available to be distributed (in which case a substantive negotiation is probably not necessary), people operating along the distributive dimension are trying to get their needs met at someone else’s expense.” (Mayer 2000:p146).
Problem-solving – Adversarial
The descriptions problem-solving and adversarial are preferred by Carrie Menkel-Meadow (Menkel-Meadow, 1984). She prefers these descriptions, as she believes the key is to look at the total conception of the negotiation rather than the tactics or strategy being used to implement that conception at any particular point in it.
Menkel-Meadow’s theory is that people get trapped in adversarial negotiation by two underlying assumptions. The first of these is that the negotiation is a zero-sum game; meaning that to the extent one side wins the other side loses. While she accepts that there can always be parts of a negotiation that are like this, adopting it as an underlying assumption seriously limits the potential for more radical solutions that might be nearer Pareto optimality (that is where the pie has been expanded to the maximum, so that all that is left is its division).
The second is the assumption that only limited items are available to be negotiated. This is derived, she believes, because negotiation is ‘in the shadow of the court’, in a phrase derived from Mnookin’s work (Cooter et al., 1982). What she means by this is that the negotiators are concentrating on the limited rulings and awards that could emanate from the court action if the negotiation fails, and so confine themselves to discussing those issues alone.
In contrast, she argues, disputants are better served by a problem-solving approach. Such an approach would involve the negotiators identifying and attempting to meet the disputants’ underlying needs and objectives (not dissimilar to interests-based negotiation). Menkel-Meadow quotes the negotiation between a husband who prefers to holiday in the mountains and a wife who prefers the seaside. An adversarial approach would result in splitting the time between sea and mountains or taking turns from year to year. A problem-solving approach would identify that the husband liked fishing and hill walking and the wife prefers swimming, sunbathing and seafood. A solution might therefore be found at a seaside resort near mountains or a mountain resort with a pool and a good restaurant with a reputation for fish.
Andrea Schneider also adopts the description of problem-solving and adversarial in her updating of the work of Williams (Schneider, 2002), further referred to below. Where she uses these terms, however, she is using them as labels for specific groups of behaviours, which is different from Menkel-Meadow’s ‘total conception of the negotiation’.
3. Effectiveness of negotiating styles
The works referred to above that differentiate between and advocate different styles of negotiation do so mainly from a theoretical perspective. Most have plenty of examples of situations where the advocated style would be likely to produce better results than the less favoured style. Examples include the well-known sisters arguing over the single orange where one wants the juice and the other the peel (quoted by Fisher and Ury, 1981, but attributed to Mary Parker Follett, Kolb, 1995), and the children arguing over chocolate cake where one wants just the icing and the other the sponge (Spencer, 2007). However few have any empirical study of the effectiveness of the styles to back up the claimed supremacy.
There are two strands of research that do this, both of which have already been mentioned. One is the study by game theorists of the choice between cooperative and non-cooperative stances in Prisoner’s Dilemma situations and the other is the research first done by Gerald Williams and followed up by Andrea Schneider.
Game theory, Prisoner’s Dilemma etc.
The Prisoner’s Dilemma concerns the position two fellow criminals find themselves in after a job together if they have agreed that they will not grass. There is enough evidence available to the police for both of them to get, say, a 1 year sentence so long as they both stick to the agreement. However if A sticks to the deal and B reneges, telling the police it was A, A would get 7 years and B would get off completely. But if both renege they would both get 4 years. What would you do? Stick to the deal and get a tolerable result, or renege in the hope of getting off completely, but only if your colleague honours the deal?
Robert Axelrod (Axelrod, 1984) argues that the most successful strategy in such cooperative/non-cooperative situations is so-called tit-for-tat. This is based on the results of the competition Axelrod organised for computer modellers to play iterative Prisoner’s Dilemma – so that participants could take into account the other player’s strategy before deciding their next move. He allocated points to the different outcomes – 3 each for both cooperating by sticking to the deal, 5 to A and 0 to B where A reneges (defects, as he calls it) and B cooperates, and 1 each where both defect.
Tit-for tat strategy, that won the competition, starts cooperatively, responds to cooperation with cooperation and responds to non-cooperation with non-cooperation. In more detail Axelrod suggested that it was important that a successful strategy should be ‘nice’, meaning that it would not be the first to be non-cooperative, ‘retaliating’ (meeting non-cooperation with non-cooperation), ‘forgiving’ (prepared to reward cooperation with cooperation, even from the previously non-cooperative) and ‘non-envious’ (that is not motivated by a desire to do better than the other party).
The best-known research study was reported on in Williams, 1983. His work was then repeated, with some amendments to the methodology, and written up in Schneider, 2002.
As mentioned, the common methodology involved several strands. Participating lawyers were asked about the lawyer with whom each had last negotiated. The respondents were asked to rate the other lawyer by reference to a list of adjectives, bipolar statements, goals and objectives and effectiveness. The adjectives included, for example, ‘accommodating’, ‘perceptive’ and ‘spineless’. Each respondent was to rate the lawyer from 0 – not at all characteristic – to 5 – highly characteristic. Bipolar statements, such as ‘took an unrealistic opening position’/‘took a realistic opening position’, and goals and objectives, such as ‘obtaining a profitable fee for self’, ‘getting a ‘fair’ settlement’ and ‘maintaining or establishing good relations between parties’ were similarly rated. Finally the opposing lawyers were rated from ineffective negotiator through average negotiator to effective negotiator.
Williams and Schneider first used statistical theory to divide the lawyers about whom they had received responses into clusters which they labelled competitive or cooperative, in the case of Williams, and adversarial or problem-solving for Schneider. They then looked at the perceived effectiveness of the lawyers in each of those clusters.
Williams’ results identified 73% of the sample as cooperative and 27% as competitive. Of the competitive negotiators 33% were seen as ineffective, 42% as average and 25% as effective. Of the cooperative negotiators, only 3% were seen as ineffective, 38% as average and 59% as effective. See figures 1 and 2 below for the spread of effective, ineffective and average negotiators amongst the competitive and cooperative negotiators respectively.
see pdf for figures
Figs. 1 and 2
Schneider took her examination further, looking at 2, 3 and 4-cluster analyses. Her 2-cluster analysis, between problem-solving and adversarial, divided the lawyers 64% problem-solving and 36% adversarial. Of the adversarial negotiators, 53% were ineffective, 37% average and 9% effective. In contrast, 54% of problem-solving negotiators were effective, 42% average and only 4% ineffective. See figures 3 and 4 below for the spread of effective, ineffective and average negotiators amongst the adversarial and problem-solving negotiators respectively.
see pdf for figures
Figs. 3 and 4
The differences between the effectiveness of the lawyers in the clusters are starkest in Schneider’s 4-cluster analysis, labelled ‘true problem-solving’ (39%), ‘cautious problem-solving’ (28%), ‘ethical adversarial’ (21%) and ‘unethical adversarial’ (12%). The unethical adversarial were 75% ineffective, 22% average and 3% effective. The ethical adversarial were 40% ineffective, 44% average and 16% effective. The cautious problem-solvers were 12% ineffective, 65% average and 24% effective. The true problem-solvers were 1% ineffective, 27% average and 72% effective. See figures 5, 6, 7 and 8 below for the spread of effective, ineffective and average negotiators amongst the unethical adversarial, ethical adversarial, cautious problem-solving and true problem-solving negotiaotiators respectively.
see pdf for figures
Figs. 5, 6, 7 and 8
4. Guidance for a person negotiating a settlement
What conclusions might someone who is about to negotiate a settlement of a dispute or disagreement draw from these works?
First, there is overwhelming academic and theoretical support, backed by research, for the proposition that the old style of competitive bargaining alone is unlikely to produce the best results. It would be a resistant negotiator who determined to follow an exclusively adversarial, distributive and position-based strategy. But it is clear that an individual’s interests cannot be served by following the opposite styles alone.
A negotiator can usefully take into account Axelrod’s conclusions of ‘nice’, ‘retaliating’, ‘forgiving’ and ‘non-envious’ as a guide from moment to moment during the course of a negotiation. However, be careful about paying too much attention to it. Axelrod’s game starts to become less clear in its guidance once it takes into account only the minor complexities of real life, such as different rounds having different levels of points allocated to them (equivalent to different negotiation points having different levels of significance) and retaliation against non-cooperative behaviour having lower significance towards the end of the game (after which the trust built up by cooperating becomes less valuable). Indeed results begin to change radically when he factors in A being unable to tell, in even as few as 10% of the rounds, whether B is being cooperative or non-cooperative, as is sometimes the case in real life.
To get the best from the suggested styles of negotiation a participant needs to compare and consider them more deeply. Although Getting to Yes is perhaps the best-known book on negotiation, it has been criticised for being naïve and self-righteous, White (1984). Even Fisher, in his comments on White’s criticism (White (1984)), acknowledges that their book is less about the world as it is and more about how intelligent people should behave. But negotiators cannot rely on their counterparts behaving as intelligent people should, in the view of Ury and Fisher, behave.
However, if you drop the label ‘principled negotiation’, with its moral tone and implication for any other style of negotiation, valuable guidance remains. The four techniques listed – ‘problem not people’, ‘interests not positions’, ‘options for mutual gain’ and ‘objective criteria’ are in themselves useful, and can be fitted into any negotiator’s toolkit.
The other two contrasting pairs, distributive/integrative and problem-solving/adversarial, have different purposes and are not just different words with similar meanings. Distributive/integrative is about the aim of the negotiation at that moment. whereas problem-solving/adversarial is about the overall style of the negotiator. So, you can easily have a problem-solving negotiator operating in the distributive phase of a negotiation or, although slightly more difficult to envisage, an adversarial negotiator in the integrative phase
No one suggests that a negotiation should always be integrative. There will be a time for integrative negotiation and one for distributive. The point is that if a negotiation moves straight to the distributive phase a huge opportunity is missed; the opportunity to increase the amount to be distributed. So those using the distributive/integrative distinction are arguing that, as a technique, any negotiator should ensure that there is an integrative phase to the negotiation.
As regards the problem-solving/adversarial distinction, a reader of Schneider’s study could argue with the labels she uses; they do not arise empirically from the study but are just chosen by her. However a negotiator can learn from the adjectives, bipolar statements and goals most prevalent in the cluster they would choose to be in. Thus, a negotiator could ask themselves would they prefer to be in cluster A, with 54% effective negotiators, 42% average and 3.5% ineffective, or cluster B, with 9% effective, 37% average and 53% ineffective.
If their answer is cluster A then they can look at the characteristics of that cluster and model their behaviour accordingly. Of course, they could choose to be in a cluster that is 72% effective, 27% average and only 1% ineffective derived from the 4-cluster analysis, labelled ‘true problem-solving’.
To be in the true problem-solving cluster a negotiator would mould their behaviour to the adjectives, bipolar statements and goals most associated with that cluster. Thus from the adjectives they would aim to be ‘ethical’, ‘trustworthy’, ‘fair-minded’, ‘dignified’ and ‘communicative’. From the bipolar statements they would be ‘courteous and honest’, ‘represent their (or their client’s) interests zealously, but within bounds’, and would ‘view the negotiation as possibly having mutual benefits’. Their goals would include ‘reaching a fair settlement that met both sides interests’ as well as ‘maximising the settlement from their own (or their client’s) point of view’.
Interested negotiators will look at the full results and use the adjectives and other traits attributed to other clusters to steer themselves away from less effective behaviours. Perhaps most significantly, they can gain considerable comfort from learning that to pursue these values, which could also be described as nobler, is also more effective than pursuing the old-fashioned aggressively adversarial approach.
But even with the empirical support for more cooperative styles of negotiating there are pitfalls. One of the most challenging aspects of negotiation is the interplay between being cooperative and competitive; dealing with the tension between creating value and claiming it. As Lax and Sebenius (Lax and Sebenius, 1986) point out, supporting the overwhelming view expressed above, there will be value-claiming in all negotiations but if negotiators only claim value, following a distributive path, joint gains will not be unearthed. Unfortunately, as they succinctly put it:
“First, tactics for claiming value…can impede its creation….
Second, approaches to creating value are vulnerable to claiming value.”
So how in practice do you find the path to the creation of the maximum value while not laying yourself open to exploitation by the tactics of the negotiator who pretends to join in value creation while in fact claiming it?
In The Manager as Negotiator Lax and Sebenius devote a chapter to this important topic. Their advice operates at a number of levels. They advocate being alert to duplicitous value creating, where the value-claimer is wearing the disguise of a value-creator and suggest tactics to head it off. One such (“petard tactics”) is, for instance, to offer a higher future payment should the seller’s suspected inflated claims for future profitability turn out to be accurate. In that way the exaggerator may find themselves trapped into accepting the contingency of the payment or revealing their lack of real belief in the claim. Another (“bait and switch”), where they are suspected to be overstating the value to them of point A in order to gain greater concessions for giving it up, is to lure them deeper into their assertion that point A is really important to them, and then give in on it against concessions in other areas.
But apart from these Lax and Sebenius support structural changes to the negotiation process. Where possible, build trust between the parties by repeated dealings. Many negotiations proceed through rounds of creating and claiming on a succession of different points. Start with smaller points where the risk, should your counterparty mislead you, is smaller, and so build up a momentum of trust. Another technique is to separate value creating from value claiming by breaking up the process into a cooperative value-creating phase and a later value-claiming phase.
More fundamentally, they see the value of involving a mediator. They identify a number of advantages: enhancing ingenuity, blunting emotional conflict escalation, facilitating information flow and enhancing communication.
5. Application to Commercial Mediation
A participant in a commercial mediation can make use of many of these conclusions. It is straightforward to be problem-solving rather than adversarial, at least in the way the term is used by Schneider, and it generally seems clear to participants to be likely to be more effective in reaching an acceptable settlement. Of course some people use adversarial tactics, even at mediation. However, in my experience, it is lawyers who have been more inclined to do so, rather than the parties themselves, and not always as part of a planned ‘good cop/bad cop’ strategy.
What is much less easy to do in a commercial mediation is genuinely to have an integrative, interests-based phase. In how many mediations do we see the parties have any real discussion beyond how strong, taking litigation risk into account, are the respective cases, what will the costs to trial be and where, given those estimates, do we think we might be able to meet?
This is, I believe, for two main reasons. First, as Lax and Sebenius point out, it is difficult to pursue the joint creation of value without opening yourself up to the possibility of being taken advantage of. Some of the tactics they advocate for use where that is suspected can work within a mediation, such as ‘petard’ and ‘bait and switch’ mentioned above. But the most effective ways of negotiators building up the trust necessary to maximise the creation of value require time and the standard mediation model, usually allocated just one day, does not give much time.
Secondly, commercial mediation is now a more or less accepted part of the litigation process. But that in itself undermines the chances of the negotiations during mediation being as effective as they could be. In the phrase adopted by Carrie Menkel-Meadow, the negotiations take place ‘in the shadow of the court’. And it is not really surprising, given that they take place between the very people who have been devoting energy and ingenuity to how to fight the litigation. It is hardly likely that those people will be able easily to swap the litigation mindset to one of finding the best settlement just because today is mediation day.
Does it matter that the standard commercial mediation model is less than perfect for ensuring more than distributive negotiation? Often, it probably does not. In many cases there is little to discuss between the parties beyond the basic how strong is the case, costs, litigation risk and can we meet somewhere? In such cases, if the mediator were to ask, what interest will be served by the position ‘pay me £1m’, the answer would be ‘the interest of getting as much out of this dispute as possible’!
But in those cases where the dispute is more complex, in those cases where commercial relationships continue, the one-day mediation model may well short change participants. For them, adopting the approach of seeing a mediation as working alongside litigation rather than just being one step in it may well be more effective. Indeed, some people even suggest having a different team trying to negotiate the settlement from the team pursuing the litigation, see, for instance Roger Fisher’s ‘Dear Lawyer’ letter, Fisher, 2002.
Perhaps in such cases mediators could be more adventurous in their suggestions for an appropriate process for reaching a settlement. Most mediators have already recognised that the mediation process can continue beyond the mediation meeting itself, but also recognise that it is a pity to miss getting to in-principle agreement while the decision-makers are engaged. Also many mediators get the process going well before the meeting with suggestions for pre-mediation ‘homework’ and telephone calls with the participants themselves and not just their lawyers.
But what about the possibility of a first meeting where the participants together discuss all the items that could go into a settlement, without at that stage assigning any relative values to them? Might this not tap into one suggested advantage that mediation has over litigation, of participants being able to design their own settlement? Given that possibility, it is perhaps a pity that, in so many mediations, discussion does not go beyond the outcomes that could be awarded by the court.
Mediators generally espouse the view that mediation is an effective way of reaching a settlement, and one in which the possibilities are greater than merely being able to choose from the elements a court could award. If so, could we not be designing models that allow for the use of the best methods of negotiation? Such models would take commercial mediation into a new phase, standing alongside litigation as a complementary conflict resolution technique.
In this article I have reviewed some of the leading work on negotiation styles and effectiveness. I have derived from this work guidance that a negotiator can take into a settlement negotiation. I have demonstrated that there is evidence to support the proposition that more principled negotiation techniques are likely to be more effective than aggressive ones.
But I have acknowledged that the path to a satisfactory agreement does not involve merely relying on cooperative techniques. A negotiator has to be prepared for the more competitive counterpart and for the value-claiming phase of any negotiation.
Finally, I have questioned whether the typical one-day commercial mediation model gives participants much opportunity to engage in integrative negotiation. I have suggested not and that, at least in some cases, this is to be regretted. I have proposed one or two ways that even the slightly adventurous mediator might look to flex the model and I have challenged the mediator community to come up with more. By doing so, as I see it, commercial mediation can truly take its place as a technique for conflict resolution rather than just a step in the litigation process.
Axelrod, R. 1984. The Evolution of Cooperation. Basic Books, New York.
Binmore, K. Review of the Complexity of Cooperation, 1 Journal of Artificial Societies and Social Simulation (1998)
Cooter, Marks & Mnookin, Bargaining in the Shadow of the Law: A Testable Model of Strategic Behavior, 11 J. Legal Stud. 225, 227 (1982).
Dresher, M. (1961). The Mathematics of Games of Strategy: Theory and Applications Englewood Cliffs, NJ: Prentice-Hall,
Fisher, R. Dear Lawyer Harvard Business Review Apr. 2002 vol. 80 issue 4 p 128
Fisher, R. and Ury, W. (1981). Getting to Yes: Negotiating agreement without giving in Boston: Houghton Mifflin
Flood, M.M. (1952). Some experimental games. Research memorandum RM-789. RAND Corporation, Santa Monica, CA.
Hoffman, D. (2003) A Primer on Successful Negotiation http://www.bostonlawcollaborative.com/blc/resources/books-articles-and-videotapes.html?branch=main&language=default Date accessed 30th November 2009
Kolb, D. M., The Love for Three Oranges, Or: What Did We Miss about Ms. Follett in the Library? 11 Negotiation J. 339, 339 (1995).
Lax, D. and Sebenius, J. (1986) The Manager as Negotiator, New York: Free Press,
Mayer, B. (2000) The Dynamics of Conflict Resolution. San Francisco: Jossey-Bass
Meltsner, M. & Schrag, P.G. (1974) Public Interest Advocacy: Materials for Clinical Legal Education Boston: Little, Brown and Company
Menkel-Meadow, C. Toward Another View of Legal Negotiation: The Structure of Problem Solving 31 UCLA L. Rev. 754 (1984)
Schneider, A.K. Shattering Negotiation Myths: Empirical Evidence on the Effectiveness of Negotiation Style 7 Harv. Negot. Rev. 143 (2002)
Spencer, D. (2007) Mediation law and practice Cambridge: Cambridge University Press
Williams, G.R. (1983), Legal Negotiation and Settlement St Paul, Minnesota: West Publishing Co.