Opinion Dynamics and Collective Decisions: Procedures, Behavior, and Systems Dynamics

我们期望民主使我们能够利用集体智慧，使我们的集体决定建立和加强社会福利，并使我们接受其分配和规范后果。集体决策应该集合我们个人的偏好和判断，但民主的三难境地是，任何集合了非二分选择的程序都必须至少违反一项民主原则：尊重多数、对多元化开放或集体理性。相反，从实证主义的角度来看，个人通过讨论和深思熟虑动态地形成和改变他们的偏好和判断。如果这些意见动力导致共识，三难困境就会得到解决，但有时意见动力会产生相反的结果：持久的两极分化意见格局。大多数集体决定和许多政治讨论都集中在对人或提案的离散选择上。这令人惊讶，因为潜在的个人判断和偏好通常是渐进的。许多决策都有选择集，这些选择集自然是连续的，如预算计划、最低收入、税率或价格、概率或风险的评估。与大多数政治模型不同，该项目关注的是决策程序，参与者可以从一个连续体中做出选择。决策程序和在意见连续体中的审议为妥协和逐步调整创造了机会。它们将被形式化为连续意见动态和集体决策的理性选择模型，其中参与者具有个人、认知和社会偏好。基于该模型，群体决策实验将有助于理解人类和群体在同时讨价还价、发现真相和协调的博弈中的行为。这将带来关于如何在特定情况下最好地解决民主三难的争论。在大群体中，意见动态自然超出了个体的范围，因此，共识、两极分化、分裂或极端主义等现象也可能是自我驱动的宏观系统动态的结果。这种宏观现象在项目中通过分析从社会调查中提取的意见景观和从互联网上获得的评级数据来量化。它们的演化通过基于主体的仿真模型来探索，这些模型通过复杂社会经济系统的物理方法被系统地分析。这有助于理解意见动态的隐藏驱动力，以及它们如何阻止或促进共识的趋同。在项目过程中，应为可能的审议程序和关于预测、税收和预算的决定的汇总得出政策建议。

Of democracy we expect that it enables us to utilize collective intelligence such that our collective decisions build and enhance social welfare, and such that we accept their distributive and normative consequences. Collective decisions shall aggregate our individual preferences and judgments, but the trilemma of democracy is that any procedure, which aggregates non-dichotomous choices, must violate at least one democratic principle: respect of majority, openness to pluralism, or collective rationality. From a positivist perspective instead, individuals dynamically form and change their preferences and judgments through discussion and deliberation. The trilemma would be solved if these opinion dynamics lead to consensus, but sometimes it that opinion dynamics produces the opposite: persistent polarized opinion landscapes. Most collective decisions and many political discussions focus on discrete choices as for people or proposals. This is surprising as the underlying individual judgments and preferences at stake are typically of a gradual nature. Many decisions have choice sets which are naturally continuous as the budget plans, the minimal income, a tax rates, or the assessment of a price, probability, or risk. In contrast to most political models, the project focusses on decision procedures where participants can make choices from a continuum. Decision procedures and deliberation in a continuum of opinion create opportunities for compromising and gradual adjustments. They will be formalized as a rational-choice model of continuous opinion dynamics and collective decision, where participants have individual, epistemic and social preferences. Based on the model, group decision experiments shall help to understand humans' and groups' behavior in such games of simultaneous bargaining, truth-finding, and coordination. This shall deliver arguments how to best solve the democratic trilemma in a specific situation. Evidence shall be found when it is good to foster and when to avoid deliberation.In large groups, opinion dynamics naturally goes beyond the scope of the individual and consequently phenomena as consensus, bipolarization, fragmentation or extremism might also be the result of self-driven macroscopic systems dynamics. Such macroscopic phenomena are quantified in the project by analysis of opinion landscapes extracted from social surveys and rating data from the Internet. Their evolution is explored by agent-based simulation models which are systematically analyzed by methods of the physics of complex socio-economic systems. This delivers an understanding of the hidden driving forces of opinion dynamics and how they prevent or foster convergence to consensus.Alongside the project, policy recommendations shall be derived for possible procedures of deliberation and the aggregation of decisions about predictions, taxation and budgeting.