Propensities to act in this or that way can be explained by the social or biological
environment and the past success of possible strategies (mutants) in it. Studying the
evolution of preferences in such a way offers the chance to combine forward-looking
deliberation (the shadow of the future focused on by most economists) and ‘path
dependence’ (the shadow of the past emphasised by biologists and psychologists). What
evolves is not behaviour itself but its determinants.
Evolution of preferences has so far been studied in a highly artificial world: a stylised
environment with only one type of interaction (a single game) determines biological or
social success. But human behaviour or, rather, its basic determinants are not gamespecific.
Men act similarly in entire, quite general classes of interaction. For example,
people usually do not mind exploiting others opportunistically in various types of markets.
But they are reluctant to do so when an experimental setting suggests a private affair (even
though it is anonymous and single-shot).
To explain how such general indications can evolve, new research by Werner Güth and
Stefan Napel, published in the October 2006 Economic Journal, studies evolution in a
stylised ‘game of life’ that encompasses different game types. The authors illustrate how
such a richer habitat can be analysed and point out how conclusions about stable
preferences depend on the compound strategic environment. In doing so, they focus on
one possible determinant of game-playing behaviour, namely other-regarding preferences
in the form of inequality aversion.
The analysis concentrates on distribution conflicts among two parties and well-known
procedures for solving them: dictatorial reward allocation and (two or three-person)
ultimatum bargaining. These game types have different ramifications concerning the
evolutionary (dis-)advantage of inequality aversion, and thus allow the investigation of very
general trade-offs concerning social preferences and the crucial issue of game specificity in
a fairly simple setting.
Depending on the relative weights of the ultimatum and the dictator game, the study
predicts either universal equity or at least partial exploitation of recipients by distributors.
The ultimatum game in isolation gives rise to very high inequality aversion and
correspondingly generous divisions of available surplus. Presence of the dictator game
may, however, substantially restrict offers in the ultimatum game. And a high frequency of
the latter can impart benevolence on dictators who in isolation will develop at most a
negligible interest in equality.
Equitable divisions are driven primarily by the fact that aversion to disadvantageous
inequality (that is, receiving less than others) results in a credible threat to reject small
offers. But analysis of the three-person ultimatum game reveals that aversion to
advantageous inequality (that is, receiving more than others) can matter, too: it is
evolutionarily beneficial to suffer from being better off than a third player because this will
extract greater material concessions from the distributor.
It is noteworthy, however, that the seemingly benevolent interest in the third party’s lot
stops exactly when the distributor starts shifting surplus to the most disadvantaged player
instead of offering the considered recipient extra compensation for subjective moral
‘suffering’.
Evolutionary studies of structurally richer habitats can yield more interesting and intuitive
results than game-specific evolution. Interestingly, the conjecture that evolutionary benefits
of non-individualistic preferences would erode as their domain is extended or as one moves
from perfect to imperfect moral discrimination between game types, is only weakly
confirmed: inequality aversion that is sufficiently strong to be noticed in ultimatum
bargaining is a fairly robust feature – not only in the experimental lab but also in
evolutionary microeconomic analysis.
ENDS
Notes for editors: ‘Inequality Aversion in a Variety Of Games – An Indirect Evolutionary
Analysis’ by Werner Güth and Stefan Napel is published in the October 2006 issue of the
Economic Journal.
The authors are in the Strategic Interaction Group at the Max Planck Institute of
Economics, Jena, Germany.
For further information: contact Werner Güth on +49-3641-686-622 (email:
karin.richter@econ.mpg.de); or Romesh Vaitilingam on 07768-661095 (email:
romesh@compuserve.com).