Dynamics of Collective Decisions

Coordination: Dr. Susanne Fuchs (HWK)


Prof. Dr. Susumu Shikano, University of Konstanz


  • Prof. Dr. André Bächtiger, University of Stuttgart
  • Prof. Dr. Martin Elff, Zeppelin University of Friedrichshafen
  • Dr. David Garcia, ETH Zürich
  • Prof. Dr. Bernhard Kittel, University of Wien
  • Dr. Jan Lorenz, Jacobs University of Bremen
  • Dr. Heiko Rauhut, ETH Zürich
  • PhD Sebastian Schutte, University of Konstanz
  • Prof. Dr. Susumu Shikano, University of Konstanz (Speaker)
  • Prof. Dr. Markus Tepe, University of Oldenburg
  • Prof. Dr. Stefan Traub, Helmut-Schmidt-University of Hamburg

Duration: December 2011 - December 2017


Conventional studies on collective decisions are usually based on game-theoretic modeling and its equilibrium solutions. While the prediction based on this theoretical foundation has been tested many times using empirical data, these studies continue to neglect the question of how the actors reach the state of equilibrium (Sober, 1983). Yet it is assumed that the social system is in a state of equilibrium at the time point of observation.

This is problematic in many respects, given the empirical analyses. First, the empirical data which do not completely correlate with the predicted equilibrium are difficult to interpret. While the similarity to equilibrium can be compared ad hoc under various empirical conditions, this has no theoretical foundation. This problem is most visible in the identification of statistical models as the zero likelihood problem (Morton, 1999). In short, theoretical models need to be able to predict all of the conditions that occur in the data (in equilibrium). Otherwise the models would be 100 percent falsified by the data, and probabilistic predictions could no longer be made. Second, the empirical evidence from cross-sectional data cannot be clearly interpreted even when it does correspond to the predicted state of equilibrium. The reason for this is the assumption that the social system at the time point of observation is in a state of equilibrium, but this cannot be confirmed by the data. Third, and most important, the implications in this approach are only derived in connection with the state of equilibrium. This means that the implications can usually only be found either in a state of equilibrium or by comparison with several equilibria. This results in the neglect of valuable information on causal dynamics in the data.

For these reasons, the study group aims to take a closer look at the dynamics of collective decisions as a process, and in particular to discuss the necessary methods. The group will address not only the dynamics of the decision-making process, but also the dynamics of repeated decisions. This does not mean that the individual decisions will be observed separately, but rather that the dependence of the decisions on each other will be considered in both temporal cross section and in time series. The dependency can be mutual mimetic contagion (Topol, 1991), bandwagon (Kandori and Rob, 1998) or another form of learning and/or adaptation process (Bendor et al., 288). These dependencies assume a form of communication in which the decision maker is informed of previous decision results and possibly over the individual decisions of the other decision makers. This makes the form and institutional condition of communication an important aspect of the study.


First, regular meetings will be held with the core members and other leading scientists in the relevant fields. Second, collaborations with local research will strengthen the regional anchoring of the HWK. Third, the study group aims to publish a book on this topic in German. There is still a large gap in this topic in the German-speaking social science community.

The first meeting of the study group took place in summer 2012.