|
|
|
XIV. 2003 IRRA BEST DISSERTATION COMPETITION
Expert Judgment in the Mediation of Collective Bargaining
Disputes
Anthony Zumbolo
New York State Public Employment Relations Board and State University
of New York, Albany
Negotiations, conflict, and the resolution
of both are studied from a variety of behavioral and social science perspectives.
Generally, each branch of learning examines these topics using tools developed
by and for researchers who share a similar viewpoint. It is infrequent
that we find the methodologies that are the common means of inquiry in
one field adopted for use in another. The practice of negotiation and
mediation is not often studied in "real-time" either, usually reenactments
or simulations are evaluated. This project strives to overcome both of
these limitations, spanning disciplines and bridging the worlds of theory
and practice.
Often the "art" of negotiating and mediating
is described through anecdotal recounting of how particular situations
are resolved. Other inquiries, e.g., game theory, take a more systematic
approach and describe the "science" of reaching agreement. This study
attempts to displace, to some degree, the "art" versus "science" debate
in favor of a more robust "art and science" evaluation of negotiation
and mediation, as suggested by Howard Raiffa (1982). This is achieved
by applying scientific methods derived from social judgment theory's examination
of the cognitive and judgmental features of the decision-making process
to active mediation cases (Mumpower et al. 1988).
Besides possessing good interpersonal communication skills, negotiators
and mediators must be adroit at interpersonal learning, that is, at gaining
knowledge from each other. The capacity to determine the degree to which
particular items are important to each of the parties in negotiations
is critical to good interpersonal learning. The essential elements of
this investigation are to ascertain how good the participants in bargaining
and mediation are at learning from each other what will resolve the conflict
and to determine if they can arrive at likely settlement terms by using
judgment theory's analytical techniques. The fact that data were collected
from the participants during the actual mediation of the negotiations
makes this study unique.
"Listening with the third ear" is essential for
a mediator to uncover the critical elements of a dispute (Kressel 1972).
Judgment theory's emphasis on the cognitive faculties connected with decision
making can help determine if the mediator's third ear is well tuned. Studying
mediation from this perspective calls for an examination of the extent
to which mediators are affected by cognitive limitations. Mediators experience
the same cognitive failings as other decision makers. Mediators have their
own unique perceptions or mind set and their capacity for managing information
is limited. The mediator's general perception of how to handle disputes,
how they should be resolved, what is important, and what is trivial, all
influence what she hears and how she interprets what is received.
Once the mediator filters what he is told by
the parties, so that it meshes with his view of the world, he begins to
make decisions about what to do with the information. At this point, just
like the negotiators, the mediator is susceptible to many cognitive failings.
Nonetheless, given these cognitive shortcomings, the sine qua
non of effective mediation is the mediator's suggesting of alternatives
for breaking deadlocks (Newman 1985). Mediators develop these important
suggestions by making judgments about the true positions of the parties.
Mediators are not insulated from cognitive foibles, and since mediator
judgments concerning suggestions are essential to the resolution of disputes,
it is salient to have a notion concerning the impact of these limitations
on mediators.
What does a mediator bring to a dispute that
helps lead to resolution? An experienced mediator may select from a number
of roles, strategies, and tactics to assist in resolving impasses (e.g.,
Kagel and Kelly 1989; Kolb 1983; Maggiolo 1985). A taxonomy of mediator
roles is introduced to help answer this question; the examples are: Reality
Messenger, Outside Reinforcer, and Interpersonal Learning Expert.
The mediator's ability to listen with the third
ear is critical to the Interpersonal Learning Expert. In this role, the
mediator develops a better understanding than the negotiators of what
is important to each side. The Expert then shines a light on or leads
the parties to the resolution that she identifies. This study focuses
not only on how well the negotiators understand each other, but also the
degree to which the mediator fulfills the role of Interpersonal Learning
Expert.
Knowing what is important to whom, i.e., interpersonal
learning, is fundamental to identifying settlement terms. Therefore, efforts
were made to select impasses for examination where the mediator would
embrace the "settlement identification model" of intervention. The guiding
precept for the "settlement identification model" is that the mediator
develop a notion of what is acceptable to the parties. He must strive
to identify likely terms of agreement.
The mediator must learn from someone what will
resolve the conflict. In the typical collective bargaining setting there
are many people from whom the mediator learns, e.g., negotiating committee
members, public officials, and the media. Not all the potential sources
of knowledge can participate in this study, hence, limited sources are
engaged. The advocates for the parties at impasse are participants in
this research and it is assumed they adequately represent what is important
to their clients. This includes incorporating elements such as political
forces, economic factors, efficiency concerns, equity considerations,
and so forth. The mediator, the advocate for the public employer, and
the chief spokesperson for the employee organization participated directly
in this research by making judgments about the acceptability of potential
contract terms.
Methods
After a specific impasse was selected,
multiple hypothetical settlements were produced for evaluation by the
mediator, the union negotiator, and the management representative. Each
of these hypothetical settlements consisted of a package of three or five
issues that were separately resolved either in favor of one of the parties'
positions or at specified intermediate points in the range between proposals.
The participants rated the acceptability of the potential contract to
each party on a scale of 1 to 20, with 20 being most acceptable. The mediator,
the union negotiator, and the management representative also evaluated
each of the package settlements based on their assessment of the likelihood
that the terms would resolve the dispute.
The potential contracts were presented to the
mediator and negotiators after the completion of the second mediation
session. Data were gathered at this point in the process to insure that
the raters would have adequate time and involvement with each other to
formulate judgments about the others' value models. This was particularly
important for the mediators because they did not have the benefit of being
involved in the direct negotiations.
Results
Responses in the judgment exercises
were examined along six dimensions. First, attention was focused on the
respondent's ability to understand what was acceptable to the employer
and the union. Participant assessment of how acceptable each potential
contract was to the union and the employer was compared to the negotiator's
own rating of acceptability. Next, the union and management negotiators'
judgment policies were externalized. These policies displayed the calculated
function forms and weights placed on each issue by the representatives.
Third, the participants' accuracy in predicting how the management and
union negotiators assigned their weights was examined. The participant's
ability to correctly rate contract terms that presented anchor and key
values was the fourth area of inquiry. The fifth area of investigation
involved predicting contract settlement terms. Finally, the predicted
settlement was compared with the contract terms that were approved by
the parties.
Participant success at predicting the acceptability
of contract terms to the negotiators was the critical element of this
analysis. Examination of correlation coefficients and the frequency distributions
of prediction errors for the participants' acceptability ratings were
the key measures for assessing such success. Results revealed that some
mediators, employers, and union negotiators were better than others at
understanding the parties' preferred contracts. From the correlation coefficient
and frequency distribution analyses, the mediators proved to be no better
at predicting acceptability of contract terms than the negotiators. This
suggests that in the impasses examined, the mediator did not hold any
privileged status with regard to interpersonal learning.
The judgment policies that negotiators and mediators rely on in assessing
the acceptability of contract terms are comprised of the weights and function
forms they attach to each issue. The weights quantify the level of importance
each issue carries for the evaluator and the function form indicates if
they prefer more or less of the item. All participants accurately identified
the appropriate function forms for both the management and union advocates.
In most instances, the weights derived from the rating exercise indicated
that the participants had a fairly good understanding of how important
each issue was in relation to all of those in dispute.
Several key indicator package contracts were
offered to evaluate the acceptability of traditional or common negotiation
resolution strategies. Primarily, the key contracts depicted classic compromising
and logrolling or horse-trading. The classic compromise offered a "half-a-loaf"
on each issue. For example, the wage adjustment that was presented was
midway between the parties' positions. In the logrolling key contracts,
issue for issue trade-offs were developed where each party found nearly
one half of its proposals and one half of their counterpart's positions
included in the potential settlement. These frequently-invoked "what if"
potential contracts did not prove to be any more acceptable to the participants
than other package proposals.
The economic technique of determining an efficient frontier was used to
evaluate how good the participants were at predicting settlement terms.
This type of analysis is useful here because it addresses the adequacy
of interpersonal learning. Do the participants know enough about each
other to recognize contract terms that are legitimate contenders, those
that the parties might actually consider for settlement? Correlation coefficients
do not necessarily tell us this. A substantial correlation coefficient
can be based on the ability to recognize potential contract terms that
the parties clearly would not accept. However, the hypothetical contracts
that fall on the efficient frontier should offer terms that the participants
believe have some degree of acceptability, thereby moderating the potential
correlation coefficient misrepresentation.
Plotting and examining the joint distribution
of the parties' preference ratings for each hypothetical contract depicted
the actual or "true" efficient frontier of settlement possibilities in
each case. To assess interpersonal learning, the predictions that the
negotiators made regarding the acceptability of the potential contracts
to their counterparts were also charted, yielding the perceived efficient
frontier. The potential contracts that fell on both efficient frontiers
were compared to the ratings they received from the mediator. What emerged
was one hypothetical contract for each case that was the predicted settlement
derived from the analysis of the judgment exercise.
The predicted settlements were compared to the
actual settlements eventually reached by the parties. Although each issue
was not resolved exactly as predicted in the judgment exercises, the actual
settlements were quite similar to those that were derived. To get a better
picture of where the final settlements landed in relation to the potential
contracts on the true efficient frontier, an estimate of the negotiators'
eventual settlement ratings was calculated from their judgment policies.
The estimates were computed by multiplying the derived judgment model
weights by the final settlement values for each issue. In each case, the
final settlement lies on or just inside the efficient frontier for potential
contracts. The proximity of the final settlements to the analytically
estimated efficient frontier validates the estimated judgment model and
settlement space.
Conclusion
This study supports the notion that during negotiation
and mediation the principal advocates and the mediator generally do a
good job of learning what will resolve an impasse. Most parties understood
their counterparts quite well. From the analysis, it is clear that the
mediator was not much better than the negotiators at predicting the acceptability
of particular contract terms. However, even when the parties understood
each other better than the mediator understood either of them, the mediator
was still able to join with the negotiators in identifying the most likely
settlement terms.
The efficient frontier analyses showed that the
participants' impressions of what would settle the impasses were fairly
accurate. They identified hypothetical contracts that might resolve the
negotiations. When those potential contracts were compared to the actual
settlement terms reached by the parties, they were not exactly the same
but quite similar.
The evidence from this study supports the proposition
that conflict, as indicated by an impasse in negotiations, can exist even
where there is good interpersonal learning by the negotiators. In these
situations, the parties may need the mediator to act as a reality messenger,
outside reinforcer, or for her to take on some role other than that of
interpersonal learning expert.
References
Kagel, S., and K. Kelly. 1989. The Anatomy of Mediation:
What Makes It Work. Washington, D.C.: BNA Books.
Kolb, D. M. 1983. The Mediators. Cambridge, Mass.: MIT Press.
Kressel, K. 1972. Labor Mediation: An Exploratory Survey. New York:
Association of Labor Mediation Agencies.
Maggiolo, W. A. 1985. Techniques of Mediation. Dobbs Ferry, N.Y.:
Oceana.
Mumpower, J. L., S. Schuman, and A. Zumbolo. 1988. "Analytical
Mediation: An Application in Collective Bargaining." In Organizational
Decision Support Systems, ed. R. Lee, A. M. McCosh, and P. Migliarese.
Amsterdam: North-Holland, pp. 61-71.
Newman, Harold R. 1985. "Mediation and Fact Finding." In The Evolving
Process--Collective Negotiations in Public Employment. Fort Washington,
Pa.: Labor Relations Press.
Raiffa, H. 1982. The Art and Science of Negotiation. Cambridge,
Mass.: Belknap/Harvard University Press.
|
