​In recent decades, society has witnessed trends towards increasing polarization, segregation and inequality. These trends can be observed in local neighborhoods (e.g. Dutch school segregation), nationally (e.g. fragmentation of the Dutch House of Representatives), and globally (e.g. income inequality). Despite enormous efforts to understand these phenomena and their contribution to social unrest and political extremism, the interdependencies between them have rarely been considered. Exploring these connections is a pressing scientific problem that can lead to breakthroughs in our understanding of society as a complex system.

Accurately modeling society in order to predict its dynamic evolution appears to be an insurmountable task, the complexity and implications of which can currently only be portrayed in science fiction, such as the “Foundation” series of books by Isaac Asimov.  These difficulties are rooted in the interconnectedness of different emergent social phenomena at different temporal and spatial scales, leading to intergenerational legal and political challenges. However, progress can be made by focusing on specific aspects of society for which the temporal and spatial scales can be widely separated. This is the case for the phenomena of polarization, segregation and inequality and their interdependencies. For example, explanations of school segregation (e.g. in the Netherlands) are rooted in an underlying inequality in educational attainment, which ultimately leads to polarized behavior, increased segregation and the promotion of further inequalities.

Tackling this challenging but feasible problem requires integrating knowledge from different disciplines into a single conceptual framework, strategically combining both academic-driven and societal (non-academic)-driven research. This is precisely the strategy of this RPA. Through interdisciplinary collaboration, RPA PSI innovatively links several strong lines of research at the UvA aimed at understanding PSI, which have so far worked in isolation, and connects to Dutch society by closely collaborating with Statistics Netherlands (CBS) and using datasets from the Municipality of Amsterdam and other sources. This directly interweaves foundational PSI methodologies including (FNWI) socio-econophysics and agent-based modelling; (FMG) formalisation of psychological theories and contextual/historical analyses; (FEB) behavioral economics and data science; (FdR) legal dogmatic/theoretical analyses and critical approaches to law; and (CBS) constraining complex models with observational data.
 
Our approach will be to gather this broad expertise and a team of postdoctoral fellows and a tenure tracker to develop a framework grounded in complexity systems theory for building models of simplified artificial societies exhibiting PSI. With this framework in hand, our goal is to address several open questions:

What are the necessary ingredients that a model must include to exhibit the emergent phenomena of polarization, segregation and inequality?
Can we use such models to develop measures/indices of societal trust? 
Are there critical levels for such indices, below which society is highly likely to suffer major societal disintegration? 
What interventions, whether political, legal or social, can regulate these emergent phenomena?

To answer these questions, we will build a computational and mathematical model of an artificial society with polarization, segregation and inequality, and constrain it using data, mostly available from CBS, with a focus both locally (Amsterdam) and nationally (Netherlands). In order to lay the foundations of the model, we need to take into account the results of contextual, historical and legal analyses. For example, current models of opinion polarization do not take into account the correlation or feedback mechanism between inequality and opinion. Projects 1, 2 and 3 therefore take societal cases that are highly relevant to the Netherlands and Amsterdam to study the mechanisms that underlie (the relationships between) polarization, segregation and inequality (see Figure 2). Parallel to this, project 4, which will run continuously over the 5-year period, uses the insights from the foundational projects to generate a computational model of PSI constrained by data from Dutch society. Specifically, these projects aim to:














(1) Uncover the reciprocal causal links between segregation and inequality . 

(2) Understand the feedback loops between inequality and polarisation . 

(3) Understand the causal links between polarisation and segregation. 

(4) Integrate (1), (2) and (3) into a single model and constrain it using available data.

You can visit our Project Hub for potential ideas for research projects within the first 3 categories above.

The main outcome of RPA PSI will be not only a fundamental understanding of these three emergent phenomena and their interrelationships but also a toolbox that can potentially be applied to understand other social phenomena (e.g. the emergence of criminal networks, the spread of misinformation in social networks, and threats to urban mental health) and societal transitions (e.g. from polarized to unpolarized societies). In addition the models can be used to study the effects of interventions, such as how urban (re)developments or reforms of housing laws can affect residential segregation in order to contribute to a more equal and fair society for future generations.