ORGANIZATIONAL BEHAVIOR AND HUMAN DECISION PROCESSES
Vol. 68, No. 1, October, pp. 68–77, 1996
ARTICLE NO. 0090
The Process of Negotiating: Strategy and Timing
as Predictors of Outcomes
MARA OLEKALNS
Department of Management and Industrial Relations, University of Melbourne, Parkville, Victoria, Australia
PHILIP L. SMITH
Department of Psychology, University of Melbourne, Parkville, Victoria, Australia
AND
THERESE WALSH
Department of Psychology, University of Melbourne, Parkville, Victoria, Australia
Using a simulated employment contract negotiation,
this research tested for differences in the use of cuing
and responding strategies across four types of out-
comes (stalemate, win–lose, suboptimum, and opti-
mum), over time, and across time and outcomes. The
use of five strategies varied over time, portraying the
negotiating process as one of decreasing flexibility. Re-
sults also showed that distributive outcomes were char-
acterized by high levels of positional information ex-
change whereas integrative agreements were charac-
terized by high levels of priority-information exchange.
It was possible to further differentiate two integrative
outcomes (suboptimum and optimum) in terms of how
two strategies were used over time: optimum outcomes
were associated with a rapid decrease in positional ar-
guing and a slight increase in restructuring, whereas
suboptimum outcomes showed the reverse pattern. Nei-
ther the frequency with which strategies were used nor
their timing was able to differentiate impasse from
win–lose outcomes. q 1996 Academic Press, Inc.
Two streams of negotiation research, both focusing
on the frequency with which particular strategies and
tactics are used, acknowledge the importance of lan-
guage in negotiation. The first stream, represented by
phase models of negotiation, has been concerned with
This research was supported by funding from a University of Mel-
bourne Special Initiatives Grant, an Economics and Commerce Fac-
ulty Research Grant and Australian Research Council Grant
SG7935603. We thank Sue Olaman and Cindy Shum for their assis-
tance in transcribing the videotapes of negotiations, and two anony-
mous reviewers for their comments on an earlier version of this pa-
per. Address correspondence and reprint requests to Mara Olekalns,
Dept. of Management & Industrial Relations, University of Mel-
bourne, Parkville, Victoria, Australia 3052.
0749-5978/96 $18.00 68
Copyright q 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.
/ a708$$2642 09-23-96 11:16:14 obhal
identifying how the distribution of tactics over time
moves negotiations to resolution (see Holmes (1992)
and Putnam (1990) for reviews). It implies that the
frequency with which tactics occur at specific times is
critical to whether negotiations will be resolved. The
second stream has focused on establishing links be-
tween the frequency with which strategies are used
and the quality of negotiated outcomes (e.g., Pruitt &
Lewis, 1975; see Putnam & Jones, 1982, for an early
review) and implies that outcome quality is primarily
a function of the proportions in which distributive and
integrative strategies are used. In this research, we
link the two approaches by considering if both fre-
quency and timing determine whether a settlement is
reached and the quality of that settlement.
A substantial proportion of the research examining
links between negotiator strategies and outcomes has
focused on establishing relationships between the fre-
quency of strategy use and the quality (integrativeness)
of negotiated outcomes. A fundamental assumption in
this research is that negotiators are strongly influenced
by their general orientation to the negotiation (e.g.,
Carnevale & Pruitt, 1992; Lewicki & Litterer, 1985;
Pruitt, 1990). Researchers have distinguished between
distributive (or competitive) and integrative (or collab-
orative) orientations and have argued that these differ-
ences in orientation result in different goals, strategies
and tactics (e.g., Putnam, 1990). There is considerable
empirical support for this argument and for the view
that these differences affect negotiator outcomes.
Negotiations that are characterized by a collabora-
tive, problem-solving orientation emphasize similari-
ties between negotiators and have as their aim the
identification of common goals (Deutsch, 1973; Lew-
icki & Litterer, 1985). They are employed when individ-
AP: OBHDP
 STRATEGIES, TIMING, AND OUTCOMES
uals place a high priority on achieving both their part-
ner’s and their own goals (Pruitt, 1983). Such negotia-
tions seek to maximize the resource pool and use
information exchange as a mechanism for establishing
trust, understanding opponents’ needs and prefer-
ences, and retaining flexibility until the close of negoti-
ations (Wilson & Putnam, 1990). Integrative outcomes
are more likely to occur when individuals adopt a col-
laborative orientation, which is characterized by more
proposals and counterproposals (Pruitt & Lewis, 1975),
problem-solving (Putnam & Wilson, 1989), discussions
about the workability of solutions (Putnam & Wilson,
1989), a greater level of agreement with opponents’
proposals (Donohue, Diez & Hamilton, 1985; Putnam &
Jones, 1989; Roloff, Tutzauer & Dailey, 1989), more
effort in uncovering underlying needs, less time in gen-
erating specific solutions (Sheppard, Blumenfeld-
Jones & Roth, 1989), and more systematic concessions
(Pruitt & Lewis, 1975). Generally, such strategies re-
flect limited positional commitment and greater flexi-
bility (Pruitt & Lewis, 1975; Sheppard et al., 1989;
Weingart et al., 1993).
Conversely, negotiations characterized by a competi-
tive orientation emphasize differences and are associ-
ated with more coercive processes aimed at influencing
opponents’ resistance points while maximizing own
gain (Deutsch, 1973; Lax & Sebenius, 1986; Lewicki &
Litterer, 1985). As a consequence, information ex-
change becomes tactical (Putnam & Jones, 1982; Wil-
son & Putnam, 1990), with negotiators being concerned
to maximize the information they receive while min-
imizing the information that they give (Walton &
McKersie, 1965). In such negotiations, the exchange of
information is used as a means to control the negotia-
tion process and is dominated by information seeking
rather than information giving (Wilson & Putnam,
1990). Increased argumentation, threats, and the as-
sertion of needs all signal an attempt to gain power
within the negotiation (Wilson & Putnam, 1990). Dis-
tributive outcomes, which are a likely consequence of
such tactics, are characterized by positional and irrele-
vant argumentation (Lewis & Fry, 1977; Lewicki &
Litterer, 1985; Pruitt & Lewis, 1975; Putnam & Wilson,
1989), personal attacks (Lewis & Fry, 1977), and high
demands and concessions (Bateman, 1980; Hamner &
Harnett, 1975; MacMurray & Lawler, 1986; Putnam &
Wilson, 1989; Smith, Pruitt & Carnevale, 1982). Fur-
thermore, increases in positional argumentation are
associated with more false information, pressure tac-
tics, and the demand for concessions (Pruitt & Lewis,
1975), as well as an increased possibility of deadlock
(Roloff et al., 1989). These tactics result in reduced in-
formation exchange, greater positional commitment
and increased hostility (Carnevale & Lawler, 1987;
/ a708$$2642 09-23-96 11:16:14 obhal
69
Weingart et al., 1993; Zubek, Pruitt, Peirce, McGillic-
cuddy & Syna, 1992).
While this research establishes strong support for a
relationship between the frequency with which negoti-
ators use distributive and integrative strategies and
the outcomes that they achieve, it treats the two styles
of bargaining and their associated outcomes as mutu-
ally exclusive. Such analyses ignore the complex,
mixed-motive nature of negotiations, which requires
individuals to both identify the available resources and
determine their distribution (e.g., Lax & Sebenius,
1986; Putnam, 1990). It further neglects the possibility
that outcomes are not determined solely by the fre-
quency with which strategies are used, but also their
timing. For example Putnam (1990), in reviewing mod-
els of negotiation, links the temporal location of integ-
rative and distributive strategies to negotiation out-
comes. On this basis, we consider whether the timing
of negotiator strategies is a useful predictor of outcome
quality.
Comparatively little research has examined whether
the distribution of strategies over time predicts either
the resolution of negotiations or the quality of that reso-
lution. Indirect support for this argument is found in
the mediation literature, which provides evidence that
tactic use across phases varies as a function of whether
or not disputes are resolved. This research has shown
that resolved and unresolved disputes differ in terms
of strategy distribution across mediation phases. Gen-
erally, agreement mediators use all tactics available to
them more evenly across the mediation session (Do-
nohue, 1989). However, they implement information
search earlier in the negotiation (Donohue, 1989) and
spend more time in the middle stages discussing facts
and interests (Donohue, Drake & Roberts, 1994) while
also managing process (Jones, 1988). By the end of ne-
gotiations, they are more focused on proposal develop-
ment (Jones, 1988). This research identifies the timing
of mediator interventions as critical in determining
whether disputes will be successfully resolved.
Further indirect support is provided by Lim and
Murnighan (1994), who tested patterns of negotiator
activity against four models. Although they found some
evidence of increases in concessions as the deadline
approached, they also reported considerable individual
differences in the frequency and pattern with which
different tactics were used. A plausible explanation is
that there is a direct relationship not only between the
frequency, but also the pattern of tactics and eventual
outcomes. Finally, Putnam, Wilson, and Turner (1990)
provide direct support by demonstrating that the tim-
ing of strategies predicts outcome quality: integrative
outcomes are associated with greater flexibility and
creativity in the final stages of negotiation, whereas
AP: OBHDP
70 OLEKALNS, SMITH, AND WALSH
distributive outcomes are associated with more posi-
tional argument in these stages.
In this research, we consider how the frequency and
timing of strategies distinguishes resolved from unre-
solved disputes, and how it affects the quality of resolu-
tion. We consider two broad classes of outcomes, dis-
tributive and integrative, and further consider the pos-
sibility of more fine-grained distinctions between two
kinds of distributive outcome (impasse and win–lose)
and two kinds of integrative outcome (suboptimum and
optimum). Win–lose outcomes are characterized by low
joint profit; suboptimum and optimum outcomes are
characterized by high joint profit and are distinguished
on the basis of their integrativeness quotient (Tripp &
Sondak, 1992).
Past research has demonstrated that distributive
and integrative outcomes are distinguished by the
more frequent use, respectively, of distributive and in-
tegrative strategies. Given the need for all negotiators
to both create and claim value (Lax & Sebenius, 1986)
we expect the two classes of outcomes to differ not only
in the frequency with which these tactics are used, but
also their distribution over time: Negotiations ending
in distributive outcomes will be dominated by the use of
distributive tactics, with integrative tactics increasing
over time to facilitate settlement; negotiations ending
in integrative outcomes will be dominated by the use
of integrative tactics, with distributive bargaining in-
creasing over time as individuals claim value.
When negotiations end in distributive outcomes, the
principal task for negotiators is to avoid a conflict spiral
that results in stalemate (Ben-Yoav & Pruitt, 1984; Pru-
itt, 1981, 1983) and this is most likely to be achieved
through the incorporation of some integrative strategies
(Pruitt, 1981). Therefore, within the class of distributive
outcomes, stalemates and win–lose outcomes will be
differentiated on the basis of the frequency of conces-
sions, which will be higher for win–lose outcomes. It is,
however, unlikely that such concessions will be evenly
spread throughout the negotiation; rather, the use of
this tactic is more likely to increase during the latter
part of the negotiation, as predicted by deadline models
(Lim & Murnighan, 1994). Two patterns should, there-
fore, distinguish win–lose outcomes from stalemates: a
decline in positional arguing over time and a comple-
mentary increase in concessions over time.
For negotiations ending in integrative outcomes the
principal task for negotiators is to avoid ending negoti-
ations before all solutions have been satisfactorily ex-
plored (Ben-Yoav & Pruitt, 1984; Pruitt, 1981, 1983;
Thompson, 1990; Weingart et al., 1993) and this is most
likely to be achieved by using a proportion of distribu-
tive strategies. Both the expansion of the resource pool
(Pruitt, 1981) and a test of opponent integrity and con-
/ a708$$2642 09-23-96 11:16:14 obhal
sistency (Wilson & Putnam, 1990) require some level
of contentiousness. Therefore, within the class of integ-
rative outcomes, the use of distributive tactics will dis-
tinguish optimum from suboptimum outcomes: Con-
tentious tactics aimed at testing the limits of the nego-
tiation should be higher for negotiations ending in
optimum agreements. Further, negotiations ending in
optimum outcomes will be characterized by a greater
use of contentious tactics early in the negotiation pro-
cess, whereas suboptimum outcomes will be character-
ized by an increasing use of contentious tactics towards
the end of negotiations.
METHOD
Subjects
One hundred twenty-eight undergraduate psychol-
ogy students at the University of Melbourne partici-
pated in the negotiation of an employment contract.
Participants were run in pairs in separate sessions.
Outcomes were classified as falling into one of four out-
comes types (stalemate, distributive, suboptimum, op-
timum) according to the criteria outlined below. Thirty-
nine outcomes were classifiable according to these cri-
teria.
Procedure
Subjects participated in the negotiation of an em-
ployment contract. Written instructions assigned the
role of either an employer or employee, described the
task, and gave them their payoff schedule and a con-
tract for recording their final outcome. They were re-
quired to reach agreement on the eight issues shown
in Table 1. As can be seen from the table, salary, annual
raise, location, and package allowed trade-offs; vaca-
tion, start date, and medical benefits were distributive,
whereas contract duration was compatible (i.e., both
negotiators earned their highest points by agreeing to
sign a 6-year contract). Subjects were instructed to
maximize the value of the contract for themselves. A
lottery offering a small cash prize was used to provide
an incentive for this; lottery tickets were allocated on
the basis of contract value. No points were awarded
for stalemates. Negotiation were conducted face-to-face
and were videotaped for subsequent analysis.
Classification of Outcomes
Outcomes were classified as follows (the number of
outcomes so classified is given in brackets). Stalemates
(9) were negotiations that were unresolved at the end of
the 25-min negotiating period. The remaining outcome
types were defined on the basis of joint outcome.
The integrativeness of outcomes was determined by
AP: OBHDP
 STRATEGIES, TIMING, AND OUTCOMES 71

TABLE 1
Payoff Matrices for Employers and Employees

Employer payoffs
Salary (Gains) Vacation (Gains) Start date (Gains) Annual raise (Gains)

$25,000 4000 2 wks 6000 1 Nov 2400 2% 1600

$26,000 3000 2 21 wks 4500 15 Oct 1800 4% 1200
$27,000 2000 3 wks 3000 1 Oct 1200 6% 800
$28,000 1000 3 21 wks 1500 15 Sept 600 8% 400
$29,000 000 4 wks 000 1 Sept 000 10% 000
Location (Gains) Contract (Gains) Package (Gains) Medical benefit

Melbourne 3200 6 years 1200 $2,000 800 25% medical 2400

Sydney 2400 5 years 900 $4,000 600 50% medical 1800
Perth 1600 4 years 600 $6,000 400 75% medical 1200
Adelaide 800 3 years 300 $8,000 300 100% medical 600
Brisbane 000 2 years 000 $10,000 000 . . . & dental 000

Employee payoffs
Salary (Gains) Vacation (Gains) Start due (Gains) Annual raise (Gains)

$25,000 000 2 wks 000 1 Nov 000 2% 000

$26,000 400 2 21 wks 1500 15 Oct 600 4% 1000
$27,000 800 3 wks 3000 1 Oct 1200 6% 2000
$28,000 1200 3 21 wks 4500 15 Sept 1800 8% 3000
$29,000 1600 4 wks 6000 1 Sept 2400 10% 4000

Location (Gains) Contract (Gains) Package (Gains) Medical benefit

Melbourne 000 2 years 000 $2,000 000 25% medical 000

Sydney 200 3 years 300 $4,000 800 50% medical 600
Perth 400 4 years 600 $6,000 1600 75% medical 1200
Adelaide 600 5 years 900 $8,000 2400 100% medical 1800
Brisbane 800 6 years 1200 $10,000 3200 . . . & dental 2400

negotiator performance on the four issues that enabled Coding of Negotiations

trade-offs. Win–lose (10) outcomes were negotiations
The coding scheme described by Donohue et al.
in which negotiator joint profit did not exceed the sum
(1981) and subsequently adapted by Alexander, Schul,
of the compromise solutions. Fully integrative agree-
and Babukus (1988) was modified to distinguish be-
ments (i.e., those in which negotiators located all poten-
tween positional and priority information exchange.
tial trade-offs) yielded a joint outcome of 14,400 points
The revised classificatory scheme allows for the coding
across the four issues; if negotiators behaved distribu-
of each utterance as one of 12 responses (to the other’s
tively, the maximum joint outcome was 9,600 points.
preceding utterance) and as one of 13 cues (for the oth-
Agreements were classified as integrative if the joint
er’s next utterance). Two raters coded all transcripts
outcome was between 12,000 and 14,400 points (12,000
and reliability was assessed using Cohen’s k. Bakeman
representing the midpoint between distributive and
and Gottman (1986) classify k in the range 0.70–0.74
fully integrative outcomes). This category was further
as very good and those at or above 0.75 as excellent.
subdivided on the basis of an integrativeness quotient
Cohen’s k was 0.74 for cues and 0.75 for responses. All
(Tripp & Sondak, 1992): A 0 to 1 scale is used to repre-
discrepancies were resolved through discussion.
sent the ratio of better to worse outcomes, with ‘‘1’’
representing settlements for which neither individual
Factor Analysis of Cues and Responses
can obtain a better outcome without worsening an op-
ponent’s outcome. Optimum solutions (12) were defined Utterances were classified as representing one of 13
as those for with in integrativeness quotient of 0.95 or cues and one of 12 responses, yielding a complex data
better. Integrative outcomes with an integrativeness set. Following Weingart, Hyder, Prietula, and Geno-
quotient below 0.95 were classified as suboptimum (8). vese (1995), the initial classification of utterances was

/ a708$$2642 09-23-96 11:16:14 obhal AP: OBHDP

72 OLEKALNS, SMITH, AND WALSH

TABLE 2
Factor Analysis of Cues

Factor

Cuing tactics Description of tactics I II III IV

Positional information Deny relevance Of an opponent’s arguments .88 .31 0.14 .11
Give positional info Arguments in support of own position .87 .06 0.001 0.01
Initial offers Makes or repeats an initial offer .85 .12 .24 .24
Treats/promises Uses threats or promises to change .79 0.06 0.21 0.12
opponent’s position
Find fault Attributes bad faith to opponent .78 .25 0.08 .26
Restructuring Restructuring Proposes a new way for proceeding 0.06 .87 .19 0.05
Openings Makes an open-ended, non-directional .25 .71 0.05 0.02
statement
Seek positional info Requests additional information .45 .65 .17 .11
about opponent’s position
Priority information Seek priority info Requests information about the value 0.15 .15 .85 0.14
of issue(s)
Give priority info Provides information about the value 0.04 .41 .75 .19
of issue(s)
Indicate flexibility Suggests a range of options or trade- .06 0.21 .73 .36
offs across issues
Concessions Request acceptance Proposes modifications to an 0.01 0.08 0.001 .91
opponent’s offer
Offer concession Makes an offer that has a lower .25 .12 .21 .77
value to self than the immediately
preceding offer
Eigen value 4.52 2.54 1.79 1.01
% variance 34.8 19.6 13.7 7.8

factor analyzed to explore the possibility that tactics
could be grouped together in a meaningful way. Subse-
quent analyses are based on the factors that emerge
from these analyses, which could be considered to rep-
resent strategies.
Factor analysis of cues. A principal components
analysis with Varimax rotation yielded four distinct
cue factors, shown in Table 2. Factor 1 includes tactics
that, according to Donohue et al.’s (1981) classificatory
scheme, represent either attacking or defending moves.
This factor includes tactics aimed at establishing and
supporting a negotiator’s position while attempting to
change an opponent’s position. These tactics are typical
of a distributive bargaining style (e.g., Lewicki & Lit-
terer, 1985) and are interpreted as providing positional
information. The second factor better represents tactics
that attempt to break negotiating scripts by either en-
gaging in nondirectional discussion or proposing a new
way of tackling the negotiation. Factor 2 is therefore
interpreted as negotiator attempts to restructure the
negotiation. Consistent with descriptions of integrative
bargaining (e.g., Lewicki & Litterer, 1985), Factor 3 is
composed of tactics that either seek or provide priority
information, or suggest tradeoffs; consequently, this
factor is interpreted as representing a priority-informa-
tion strategy. Finally, the two tactics of proposing mod-
ifications to an opponent’s offer or making concessions
lead us to interpret Factor 4 as representing conces-
sionary behavior.
Factor analysis of responses. A principal compo-
nents analysis with Varimax rotation yielded three dis-
tinct response factors, shown in Table 3. Two of the fac-
tors that emerged in this analysis again fit the descrip-
tions of distributive and integrative bargaining. Factor
1 is represented by attacking and defending tactics that
focus on opponents’ positional arguments, coupled with
contentious tactics aimed at changing that position. This
factor is interpreted as representing a positional infor-
mation strategy. Factor 2, which is interpreted as repre-
senting a priority information strategy, is composed of
what Donohue et al. (1981) classify as integrative tac-
tics: giving and seeking priority information, as well as
a willingness to accept offers. Finally, the third factor
also represents a contentious style of bargaining in
which negotiators either reject opponents’ offers or re-
state their demands. This factor is interpreted as rejec-
tion, as these tactics reject opponents’ positions either
explicitly or implicitly. It will be noted that one response
tactic, accepts arguments, has a negative loading on the

/ a708$$2642 09-23-96 11:16:14 obhal AP: OBHDP

 STRATEGIES, TIMING, AND OUTCOMES 73

TABLE 3
Factor Analysis of Responses

Factor

Responding tactics Description of tactics I II III

Positional information Give positional info Arguments in support of own position .88 0.05 0.05
Seek positional info Requests additional information about opponent’s .85 0.25 0.14
position
Reject argument Opponent’s arguments are treated as irrelevant to .85 0.19 .23
settlement
Continue discussion . . . Of previous topic .82 .28 .03
Personal insult Rejects opponent’s argument with a personal insult .63 0.03 .54
Topic change Introduces a new topic for discussion .66 .23 .13
Priority information Give priority info Provides information about the value of issue(s) .07 .92 0.11
Seek priority info Requests information about the value of issue(s) .06 .88 .04
Accept offer Explicit acceptance of opponent’s proposal 0.09 .55 .28
Rejection Reject offer Explicit rejection of opponent’s offer .48 .08 .71
Assert wants States requirements and expects compliance .55 .22 .56
Accept argument Accepts opponent’s point of view, but not offer .24 0.02 0.72
Eigen value 4.72 2.29 1.43
% variance 39.3 19.1 12.0

third response factor. Consequently, it is retained as a VAs, one for cues and one for responses, were con-
single tactic in further analyses. ducted.2
The preceding discussion of tactics and outcomes
RESULTS suggested (a) that tactical differences would distin-
guish integrative from distributive outcomes; (b) that
Negotiations ranged in length from 10 to 25 min, optimum integrative outcomes would be distinguished
consequently differing in the total number of strategies from suboptimum ones; (c) that win–lose outcomes
used. This raises the possibility that any differences would be distinguished from stalemates. A series of
that emerged if analyses were conducted on raw fre- three planned comparisons on the Outcome factor was
quencies might simply reflect these differences in over- used to test these hypotheses: Contrast 1 compared
all length. Because our hypotheses involve the relative, distributive and integrative outcomes, Contrast 2 com-
rather than the absolute, frequency with which partic- pared optimum and suboptimum outcomes, and Con-
ular strategies are used, the analysis was carried out trast 3 compared stalemates and win–lose outcomes.
on the proportion of times a particular strategy was We argued that the use of strategies and tactics could
used by a dyad during a given phase of the negotiation. differ over time; we further argued that trends over
To avoid problems of linear dependency between values time may serve to differentiate between the four cate-
of the dependent variable (because the sets of propor- gories of outcomes. To test these hypotheses, each tran-
tions always sum to one), one of the strategies was script was divided into six equal segments, determined
arbitrarily dropped from the analysis to create a lin- by the number of utterances. To test for differences
early independent set, in the manner described by Har- in the distribution of tactics over time, the linear and
ris (1985, pp. 109–115). Before carrying out the analy- quadratic trend components were also extracted.
sis, the remaining proportions were subject to a 2arc-
q
sin p transformation to stabilize the variances, as Analysis of Cuing Strategies
described by Winer (1971, pp. 399–400).1 The trans-
The relationship between Time, Outcome and Cuing
formed proportions were then subject to a multivariate
Strategy was examined using MANOVA, with the pro-
analysis of variance (MANOVA). Two separate MANO-
portion of strategies within each time segment provid-
ing the dependent variable. Four significant interac-
1
Because there are only n 0 1 degrees of freedom in a set of n tions were found: Strategy interacted with Outcome
proportions, this procedure does not result in any loss of information.
In the subsequent analysis of the simple effects of tactic type, how-
2
ever, all n tactics were analyzed so that their effects over time could As neither analysis violated sphericity assumptions, univariate
be characterized individually. results are reported (Tabachnik & Fidell, 1983).

/ a708$$2642 09-23-96 11:16:14 obhal AP: OBHDP

74 OLEKALNS, SMITH, AND WALSH
FIG. 1. Mean frequency of cuing strategy use over time.
(distributive vs integrative), F(2,70) Å 5.97, p õ .005;
with Outcome (suboptimum vs optimum), F(2,70) Å
3.33, p õ .05; and with Time, F(10,350) Å 3.09, p õ
.001. These last two interactions are qualified by a
three-way interaction between Strategy, Time and
Outcome (suboptimum vs optimum), F(10,350) Å 2.41,
p õ .01. To further explore these effects, the interac-
tions were decomposed by strategy into their simple
effects.
Analysis of simple effects showed, first, that two cu-
ing strategies varied in their use over time. As can be
seen in Fig. 1, a significant quadratic trend in the use
of positional-information strategies, F(1,175) Å 4.21, p
õ .05, showed this strategy to peak in the middle stages
of the negotiation. Conversely, a quadratic trend in the
use of concession-exchange strategies, F(1,175) Å 8.47,
FIG. 2. Mean frequency of cuing strategy use as a function of outcomes and time.
/ a708$$2642 09-23-96 11:16:14 obhal
p õ .005, was lowest in the middle stages of the negotia-
tion.
The use of two cuing strategies differentiated distrib-
utive and integrative outcomes. When negotiations
ended in distributive outcomes, dyads used more posi-
tional-information strategies, F(1,35) Å 15.54, p õ
.001, and fewer priority-information strategies, F(1,35)
Å 15.54, p õ .001, than when negotiations ended in
integrative outcomes. The more frequent use of posi-
tional-information strategies, F(1,35) Å 5.23, p õ .05,
also differentiated suboptimum from optimum out-
comes, although this finding was qualified by a three-
way interaction with Time, as described below.
Within the class of integrative outcomes, linear
trends in the use of positional-information strategies,
F(1,175) Å 9.75, p õ .005, and restructuring strategies,
F(1,175) Å 12.72, p õ .001, over time further distin-
guished optimum and suboptimum outcomes. As can
be seen in Fig. 2, the use of positional-information
strategies increased marginally over time when negoti-
ations ended in suboptimum outcomes and decreased
over time when negotiations ended in optimum out-
comes. Conversely, the use of restructuring strategies
decreased over time when negotiations end in subopti-
mum outcomes, and increased marginally over time
when they ended in optimum outcomes. It is also evi-
dent that, over time, the use of these strategies di-
verged when negotiations ended in suboptimum out-
comes, converging when they ended in optimum out-
comes.
Analysis of Responding Strategies
The relationship among Time, Outcome, and Re-
sponding Strategy was examined using MANOVA,
with the proportion of strategies within each time seg-
ment providing the dependent variable. Two significant
AP: OBHDP
 STRATEGIES, TIMING, AND OUTCOMES
FIG. 3. Mean frequency of responding strategy use over time.
interactions emerged. Strategy interacted with both
Outcome (distributive vs integrative), F(2,70) Å 3.69,
p õ .05, and Time, F(10,350) Å 2.41, p õ .01. To further
explore these effects, the interactions were decomposed
by strategy into their simple effects.
Analysis of simple effects showed that three response
strategies varied in their use over time. These trends
are shown in Fig. 3. A significant quadratic trend in
the use of positional-information responses, F(1,175) Å
14.06, p õ .001, showed this strategy to peak in the
middle stages of the negotiation. Conversely, a qua-
dratic trend in the use of rejection responses, F(1,175)
Å 8.32, p õ .005, was lowest in the middle stages of
the negotiation. Finally, a linear trend, F(1,175) Å 6.83,
p Å .01, showed the acceptance of opponent’s argu-
ments to decrease over time.
The use of one response strategy differentiated dis-
tributive and integrative outcomes. In a pattern paral-
leling findings for the use of cue strategies, when nego-
tiations ended in distributive outcomes, dyads used
fewer priority-information strategies, F(1,35) Å 8.62,
p õ .01, than when negotiations ended in integrative
outcomes. A second strategy, the use of position infor-
mation, was a marginally significant differentiator,
F(1,35) Å 3.83, p Å .058, between distributive and in-
tegrative outcomes.
DISCUSSION
Several strategies varied over time independent of
the way in which negotiations ended. Not only are the
temporal patterns broadly consistent with phase mod-
els of negotiation (as described by Holmes, 1992), they
also portrayed the negotiating process as one of de-
creasing flexibility. First, the more frequent use of con-
cessionary cues and rejection responses early and late
/ a708$$2642 09-23-96 11:16:14 obhal
75
in negotiations suggests that their role varies as a func-
tion of temporal context: They signal flexibility, test
firmness and define the bargaining zone early in the
negotiations while late in the negotiations they are
used to reach settlement. (Note that this is consistent
with the arguments made by several authors [Lewis &
Fry, 1977; Putnam, 1990; Putnam & Wilson, 1989;
Roloff et al., 1989] that the same strategy may serve
either integrative or distributive functions, depending
on context.) Second, negotiators become progressively
less willing to respond by accepting opponent’s argu-
ments as the negotiation progresses. Thus, the limits
of the negotiation are defined early in the negotiation
process. The key means for challenging these limits is
in the high levels of positional-information cuing and
responding cuing and responding during the middle
phases of the negotiation.
Consistent with past research, our results showed
that integrative and distributive outcomes were associ-
ated with different levels of positional and priority in-
formation exchange. Whereas distributive outcomes
were characterized by a higher proportion of positional-
information cues and responses, as well as a lower pro-
portion of priority-information cues, the reverse was
true for integrative outcomes. These results clearly fit
with Fisher and Ury’s (1981) distinction between posi-
tional and principled bargaining and further demon-
strate that low levels of positional arguing are associ-
ated with functional outcomes. They are also consistent
with past research showing that integrative outcomes
are characterized by higher levels of information ex-
change (Thompson, 1990; Thompson & Hastie, 1990).
The frequency with which tactics are used did not, how-
ever, enable more fine-grained distinctions within the
two outcome types to be made. This suggests that, al-
though the frequency with which strategies are used
AP: OBHDP
76 OLEKALNS, SMITH, AND WALSH
channels negotiators to either a distributive or an in-
tegrative path, it does not determine the quality of out-
comes within those paths.
Extending this point, negotiations that ended in opti-
mum and suboptimum outcomes differed in the use of
two cuing strategies (positional information; restructur-
ing) over time. Two patterns differentiated optimum from
suboptimum outcomes: The relative stability of these
strategies over time and the extent to which their use
had converged or diverged by the end of negotiations.
Putnam (1994) argues that the way in which negotiators
respond to their partners’ cues conveys implicit messages
about how negotiations should be interpreted and struc-
tured. We propose that how cues are distributed over
time also conveys an implicit message, signalling a nego-
tiators’ dominant motive. Because its use is more predict-
able, the more stable of the two strategies establishes
the negotiating context; use of the less stable strategy is
interpreted within this context.
Earlier, we argued that negotiators must both create
and claim value to optimize outcomes. For example,
Pinkley, Griffith, and Northcraft (1995) demonstrated
that the self-reported use of mixed-motive strategies
is associated with better joint outcomes. The second
pattern described above identifies the management of
these strategies as critical to achieving optimum out-
comes. Achieving optimum outcomes is a cognitively
complex task that requires dyads to both differentiate,
recognizing multiple perspectives on a problem, and
integrate, building relationships between these per-
spectives (definitions as given by Gruenfeld, 1995). In
the present case, positional information seems better
suited to differentiation and restructuring better suited
to integration. Our results then suggest that when
these tasks are accomplished affects outcome opti-
mality. Suboptimal negotiators focus increasingly on
the task of differentiation over time, whereas optimum
negotiators focus increasingly on the task of integra-
tion over time. In doing so, optimum negotiators estab-
lish a balance between the two strategies suggesting
that this, too, is critical to the quality of integrative
outcomes.
The distribution of strategies over time did not, how-
ever, differentiate stalemates from win–lose outcomes.
Information processing problems and the cognitive rep-
resentation of negotiations provide alternative expla-
nations for outcome differences in these dyads. Pinkley
et al. (1995) found that fixed-pie expectations affect
both information availability (by affecting search strat-
egies) and information processing, consequently lead-
ing to low joint outcomes. Although our results rule out
the possibility of differences in information availabil-
ity, they do not rule out differences in information pro-
cessing. A second difference between these dyads may
/ a708$$2642 09-23-96 11:16:14 obhal
be the ability to define the negotiation in common
terms. Pinkley and Northcraft (1994) have shown that
representations of conflict converge over time and spec-
ulate that this may be a necessary condition for resolu-
tion. On this basis, we propose that impasse dyads fail
to develop a common understanding of the problem,
and further speculate that this may, in turn, affect
information processing. Why such differences occur in
distributive negotiations is an open question. At a
macro-level, situational and dispositional factors may
impact on cognitive representations of conflict. At a
micro-level, it may be the specific sequencing of strate-
gies (rather than their frequency or temporal distribu-
tion) that facilitates or inhibits the development of
shared meaning.
We were able to distinguish distributive and integ-
rative outcomes in terms of the frequency of strategy
use, suggesting that information availability or suc-
cessful differentiation moves negotiators from distribu-
tive to integrative outcomes: priority information ex-
change and negotiator flexibility provide the informa-
tion necessary for recognizing differing priorities.
However, maximizing benefits also requires effective
information processing or integration. This is achieved
when negotiators are able to restructure the negotiat-
ing process, achieving a balance between conten-
tiousness and cooperation. Our results suggest that,
whereas the frequency with which priority and posi-
tional-information strategies (as both cues and re-
sponses) are used is sufficient to move negotiators from
distributive to integrative outcomes, the temporal dis-
tribution of two cuing strategies (positional informa-
tion and restructuring) is critical to whether dyads
fully integrate.
REFERENCES
Alexander, J. F., Schul, P. L., & Babakus, E. (1991). Analyzing inter-
personal communications in industrial marketing negotiations.
Journal of the Academy of Marketing Sciences, 19, 129–139.
Bateman, T. S. (1980). Contingent concession strategies in bar-
gaining. Organizational Behavior and Human Performance, 26,
212–221.
Bakeman, R., & Gottman, J. M. (1986). Observing interaction: An
introduction to sequential analysis, Cambridge: Cambridge Univ.
Press.
Ben-Yoav, O., & Pruitt, D. G. (1984). Resistance to yielding and the
expectation of cooperative future interaction. Journal of Experi-
mental Social Psychology, 20, 323–335.
Carnevale, P. J. D., & Lawler, E. J. (1987). Time pressure and the
development of integrative agreements in bilateral negotiations.
Journal of Conflict Resolution, 30, 636–659.
Carnevale, P. J., & Pruitt, D. G. (1992). Negotiation and mediation.
Annual Review of Psychology, 43, 531–582.
Deutsch, M. (1973). The resolution of conflict. New Haven: Yale Univ.
Press.
Donohue, W. A. (1989). Communicative competence in mediators. In
AP: OBHDP
 STRATEGIES, TIMING, AND OUTCOMES
K. Kressel, D. G. Pruitt and Associates (Eds.), Mediation Research:
The Process and Effectiveness of Third-Party Intervention. San
Francisco: Jossey-Bass Publishers.
Donohue, W. A., Drake, L., & Roberts, A. J. (1994). Mediator issue
intervention strategies: A replication and some conclusions. Medi-
ation Quarterly, 11, 261–274.
Donohue, W. A., Diez, M. E., & Hamilton, M. (1984). Coding natural-
istic negotiation interaction. Human Communication Research, 10,
403–425.
Fisher, R., & Ury, W. (1981). Getting to Yes: Negotiating Agreement
Without Giving In. Great Britain: Arrow Press.
Gruenfeld, D. H. (1995). Status, ideology and integrative complexity
on the U.S. Supreme Court: Rethinking the politics of political
decision making. Journal of Personality and Social Psychology, 68,
5–20.
Hamner, W. C., & Harnett, D. L. (1975). The effects of information
and aspiration level on bargaining behavior. Journal of Experimen-
tal Social Psychology, 11, 329–342.
Harris, R. J. (1985). A Primer of Multivariate Statistics (2nd ed.).
Orlando, FL: Academic Press.
Holmes, M. E. (1992). Phase structures in negotiation. In L. L. Put-
nam, & M. E. Roloff (Eds.), Communication and Negotiation, New-
bury Park: Sage Publications.
Jones, T. S. (1988). Phase structures in agreement and no-agreement
mediation. Communication Research, 15, 470–495.
Lax, D., & Sebenius, J. (1986). The Manager as Negotiator. New
York: Free Press.
Lewicki, R. J., & Litterer, J. A. (1985). Negotiation. Homewood, IL:
Irwin.
Lewis, S. A., & Fry, W. R. (1977). Effects of visual access and orienta-
tion on the discovery of integrative bargaining alternatives. Orga-
nizational Behavior and Human Performance, 20, 75–92.
Lim, S. G-S., & Murnighan, J. K. (1994). Phases, deadlines and the
bargaining process. Organizational Behavior and Human Decision
Processes, 58, 153–171.
MacMurray, B. K., & Lawler, E. J. (1986). Level-of-aspiration theory
and initial stance in bargaining. Representative Research in Social
Psychology, 16, 35–44.
Pinkley, R. L., & Northcraft, G. B. (1994). Conflict frames of refer-
ence: Implications for dispute resolution and outcomes. Academy
of Management Journal, 37, 193–205.
Pinkley, R. L., Griffith, R. L., & Northcraft, G. B. (1995). ‘‘Fixed Pie’’
a la mode: Information availability, information processing and
the negotiation of suboptimal agreements. Organizational Behav-
ior and Human Decision Processes, 62, 101–112.
Pruitt, D. G. (1990). Problem solving and cognitive bias in negotia-
tion: A commentary. In B. H. Sheppard, M. H. Bazerman, & R. J.
Lewicki (Eds.), Research on Negotiation in Organizations (Vol. 2).
Greenwich, CT: JAI Press Inc.
Pruitt, D. G. (1983). Strategic choice in negotiation. American Behav-
ioral Scientist, 27, 167–194.
Pruitt, D. G. (1981). Negotiation Behavior. New York: Academic
Press, Ltd.
Pruitt, D. G., & Lewis, S. A. (1975). Development of integrative solu-
tions in bilateral negotiations. Journal of Personality and Social
Psychology, 31, 621–630.
Putnam, L. L. (1994). Productive conflict: Negotiation as implicit
Received: June 26, 1996
/ a708$$2642 09-23-96 11:16:14 obhal
77
coordination. The International Journal of Conflict Resolution, 5,
284–298.
Putnam, L. L. (1990). Reframing integrative and distributive bar-
gaining: A process perspective. Research on Negotiation in Organi-
zations, 2, 3–30.
Putnam, L. L., Wilson, S. R., & Turner, D. B. (1990). The evolution
of policy arguments in teachers’ negotiations. Argumentation, 4,
129–152.
Putnam, L. L., & Jones, T. S. (1982) Reciprocity in negotiations: An
analysis of bargaining interaction, Communication Monographs,
49, 171–191.
Putnam, L. L., & Wilson, S. R. (1989). Argumentation and bargaining
strategies as discriminators of integrative outcomes. In M. A.
Rahim (Ed.), Managing Conflict: An Interdisciplinary Approach,
New York: Praeger.
Putnam, L. L., & Jones, T. S. (1982). Reciprocity in negotiations: An
analysis of bargaining interaction. Communication Monographs,
49, 171–191.
Roloff, M. E., Tutzauer, F. E., & Dailey, W. O. (1989). The role of
argumentation in distributive and integrative bargaining contexts:
seeking relative advantage but at what cost? In M. A. Rahim (Ed.),
Managing Conflict: An Interdisciplinary Approach, New York:
Praeger.
Sheppard, B. H., Blumenfeld-Jones, K., & Roth, J. (1989). Informal
thirdpartyship: Studies of everyday conflict intervention. In K.
Kressel, D. G. Pruitt and Associates (Eds.), Mediation Research:
The Process and Effectiveness of Third-Party Intervention. San
Francisco: Jossey-Bass Publishers.
Smith, D. L., Pruitt, D. G., & Carnevale, P. J. D. (1982). Matching
and mismatching: The effect of own limit, others toughness and
time pressure on concession rate in negotiation. Journal of Person-
ality and Social Psychology, 42, 876–883.
Tabachnik, B. G., & Fidell, L. S. (1983). Using Multivariate Statistics.
New York: Harper & Row, Publishers.
Thompson, L. (1990). An examination of experienced and naive nego-
tiators. Journal of Personality and Social Psychology, 59, 82–90.
Thompson, L., & Hastie, R. (1990). Social perception in negotiation.
Organizational Behavior and Human Decision Processes, 47, 98–123.
Tripp, T. M., & Sondak, H. (1992). An evaluation of dependent vari-
ables in experimental negotiation studies: Impasse rates and par-
eto efficiency. Organizational Behavior and Human Decision Pro-
cesses, 51, 273–295.
Walton, R. E., & McKersie, R. B. (1965). A Behavioral Theory of
Labor Negotiations: An Analysis of a Social Interaction System,
Ithaca: NY: ILR Press.
Weingart, L. R., Prietula, M. J., Hyder, E. B., & Genovese, C. R. (1995).
Tactical knowledge and the process of negotiation. Paper presented
at the Academy of Management Meeting, Vancouver, BC.
Weingart, L. R., Bennett, R. J., & Brett, J. M. (1993). The impact
of consideration of issues and motivational orientation in group
negotiation process and outcome. Journal of Applied Psychology,
78, 504–517.
Wilson, S. R., & Putnam, L. L. (1990). Interaction goals in negotia-
tion. Communication Yearbook, 13, 374–406.
Winer, B. J. (1971). Statistical Principles in Experimental Design
(2nd ed.). Tokyo, Japan: McGraw-Hill Kogakusha.
Zubek, J. M., Pruitt, D. G., Peirce, R. S., McGillicuddy, N. B., & Syna,
H. (1992). Disputant and mediator behaviors affecting short-term
success in mediation. Journal of Conflict Resolution, 36, 546–572.
AP: OBHDP