Prospect Theory, Fairness, and the Escalation of Conflict at a Negotiation Impasse*

AuthorTopi Miettinen, Olli Ropponen, Pekka Sääskilahti
DOIhttp://doi.org/10.1111/sjoe.12384
Publication Date01 Oct 2020
Scand. J. of Economics 122(4), 1535–1574, 2020
DOI: 10.1111/sjoe.12384
Prospect Theory, Fairness, and the
Escalation of Conflict at a Negotiation
Impasse*
Topi Miettinen
Hanken School of Economics, FI-00101 Helsinki, Finland
topi.miettinen@hanken.fi
Olli Ropponen
VATT Institute for Economic Research, FI-00101 Helsinki, Finland
olli.ropponen@vatt.fi
Pekka S¨askilahti
Compass Lexecon, FI-00100 Helsinki, Finland
psaaskilahti@compasslexecon.com
Abstract
We study a bilateral negotiation set-up where,at a bargaining impasse, the disadvantaged party
chooses whether to escalate the conflict or not. Escalation is costly for both parties, and it results
in a random draw of the winner of the escalated conflict. We derive the behavioral predictions of
a simple social utility function, which is convex in disadvantageous inequality, thus connecting
the inequity aversion and the prospect theory models. Our causal laboratory evidence is, to a
large extent, consistent with the predicted effects. Among other things, the model predicts that
the escalation rate is higher when escalation outcomes are riskier, and that the disagreement rate
is lower when the cost of escalating the conflict is higher.
Keywords: Bargaining; conflict; inequity aversion; loss aversion;quantal response equilibrium
JEL classification:C72; C91; D03
I. Introduction
Wage negotiations between a cor porate employer and a labor union can stall
and escalate into a strike, which harms both parties. Settlement negotiations
*The views presented do not necessarily represent those of our employers. We thank C. G¨oring,
T. akel ¨a, and M. Ploner for research assistance. We also thank G. Bolton, W. G¨uth, A. Isoni,
K. Kotakorpi, M. Liski, P. H. Matthews, T. Nurminen, M. Ploner, O. Rydval, A. Lindholm,
and seminar audiences at NHH, Helsinki GSE, IFN, Innsbruck, Jena, EEA-ESEM, and BEET
2016 at Bilkent University for insightful comments.We acknowledge the financial support of the
NorwegianResearch Council (250506), the IPR University Center, the Yrj¨oJahnsson Foundation
(7011), and the Max Planck Institute of Economics.
Also affiliated with Helsinki Graduate School of Economics.
C
The editors of The Scandinavian Journal of Economics 2019.
1536 Prospect theory and escalation of conflict
between a plaintiff and a defendant of a lawsuit can stall, and the plaintiff
might take the case to court, which can entail high legal expenses for both
parties and uncertainties about the final verdict. Negotiations in a territorial
conflict can stall and trigger armed conflict, with devastating consequences.
There have been more than 10 million battle casualties worldwide since
World War II (Lacina and Gleditsch, 2005). In 2000, in US State courts
alone, about 20 million cases were filed, of which about 3–4 percent ended
up in trial. This leaves the courts with a workload of about a million cases
annually (Ostrom et al., 2003). Strikes and labor unrest have a negative
effect on productivity and product quality (Krueger and Mas, 2004; Mas,
2008), and Gruber and Kleiner (2012) have shown that nurses’ strikes
increased in-hospital mortality by 18.3 percent in the state of New York.
As noted by Fearon (1995) and Jackson and Morelli (2011), the failure of
bargaining is a key prerequisite for all of these inefficiencies to arise.
In this paper, we design a simple non-framed experiment to better
understand how inefficient conflict comes about between two individuals,
and how the disadvantaged dispute party might engage in an escalation
of the conflict when negotiations stall. We are interested in testing the
implications of a social utility model that makes explicit the connection
between the prospect theory value function (Kahneman and Tversky, 1979;
Tversky and Kahneman, 1992) and the inequity aversion model of Fehr
and Schmidt (1999). We hypothesize that, at a negotiation impasse, social
comparison importantly influences the decision whether to escalate conflict
or not, and that the convexity of the social utility in disadvantageous pay-off
inequality is the key to understand these decisions. These hypotheses are
captured in an “inequity-as-loss” utility function that we propose. This form
of the social utility function was suggested by Loewenstein et al. (1989),
who found that disadvantageous inequality can be accounted as a loss in
the prospect theory sense. The successful social preference literature that
followed (Bolton, 1991; Fehr and Schmidt, 1999; Bolton and Ockenfels,
2000; Charness and Rabin, 2002) paid less attention to the curvature
properties and focused more on the first-order effects of inequity aversion
and fairness. If disadvantageous inequality is perceived as a loss in this
manner, then the implications could be dramatic for settlement patterns: the
riskier the escalation outcomes are, the more the disadvantaged provokers
will escalate conflict – not less, as suggested by risk aversion.1
1In addition to social comparison, high loss references can be driven, for instance, by high
aspirations set at the negotiation table (Korobkin, 2002; Karag¨ozoglu and Keskin, 2018). See
also Cox et al. (2007) and Bolton and Ockenfels (2000) for models of non-linearities in inequity
aversion. Eisenkopf and Teyssier (2013) study the effect of envy and loss aversion in contests,
but without convexity effects.
C
The editors of The Scandinavian Journal of Economics 2019.
T. Miettinen, O. Ropponen, and P. S¨askilahti 1537
From an applied perspective, there is a pressing need for understanding
such effects because of their potential effects on such important frictions as
labor disputes or the burden of courts. More generally, better joint models
of risk and other-regarding preference hold a promise of yielding higher
explanatory power in any strategic interaction context.
In our experimental design, parties first attempt a settlement. A failure
to strike a deal puts one of the parties at a disadvantaged position, in
the sense that their pecuniary pay-off is lower than that of the opposing
side at all ensuing conflict outcomes. This underdog is given an option
either to acquiesce or to engage in inefficient rent-seeking (i.e., to escalate
conflict in order to potentially reach a still disadvantaged but less unequal
outcome). The decision to escalate results in a lottery with an exogenous
and publicly known probability of winning and losing, and equally large
publicly known expenses to each side of the dispute. We experimentally
vary the following: (i) the probability of winning of the disadvantaged party;
(ii) the cost of escalation of conflict; (iii) while preserving the expected
pay-offs at escalation, whether the escalation outcome is risky or certain.
In our set-up, a lower probability of winning and higher escalation costs
are perfect substitutes in reducing escalation incentives for a risk-neutral
and self-interested underdog.
Regarding escalation, we observe the following. (i) An escalator’s lower
probability of winning and (ii) higher costs of escalation both reduce
the escalation rate. We also observe, going against risk-aversion, that
(iii) greater variation in escalation outcomes increases the escalation rate.
Among the conditions where rational self-interest predicts no escalation, the
observed escalation rate is highest when escalation is costly and offers a
high chance of rendering pay-offs more equal. In fact, with risky outcomes,
escalation is more frequent than refraining from it, even if doing so
is suboptimal from a self-interested perspective. Regarding negotiations
preceding the escalation choices, we find that settlement rates are highest
when it is expensive to escalate conflict. With high costs, escalation threat
works by making the negotiators more careful in seeking advantage in
settlement.
The observed escalation patterns are, to a large extent, in line with
the predictions of the proposed inequity-as-loss model. The model even
explains why lowering the underdog’s probability of winning and making
escalation outcomes less risky curb inefficient escalation more effectively
than increasing the costs of escalation and making the outcomes more
risky.2
2Inspired by the ideas of Veblen (1899), Robson (1992) theoretically studies the effect of status
concerns on risk-taking, and shows that utility can become convexin wealth as a result of indirect
C
The editors of The Scandinavian Journal of Economics 2019.

To continue reading

Request your trial