Exploring the “spark of life” moment when individuals form institutions
By Milton J. Friesen and Srikanth Mudigonda
While computers and instantaneous communications seem to have increased the complications of our daily lives, in the hands of researchers, these tools can move us toward a form of simplicity that furthers understanding of complex dynamics.
One of the tools available to researchers are computer models that take often-difficult and complex phenomena and usefully simplify those dynamics so that we can see how things work. In our work, we are trying to gain a better understanding of how social and economic inequality work by paring down the dynamics to their most simple form using a computer model populated by agents. Agents are computational objects that represent key cognitive and decision-making aspects of humans. Agents' actions in a model represent the corresponding human actions in the phenomenon the model aims to simulate. These agents can relate to other agents and cooperate in simple ways. How does cooperation, pooling resources with other agents, improve their ability to survive and gain resources? How might non-cooperation lead to a decrease in the odds of survival for agents in their environment?
These are, of course, long-standing and vital human and social questions that we have, so far, not fully answered. In our paper “Institutional Emergence and the Persistence of Inequality in Hamilton, ON 1851-1861” we make use of insights that urban scholar Michael Katz made in his landmark mid-1970s study (The People of Hamilton Canada West: Family and Class in a Mid 19th Century City) on structural inequality to develop an agent-based model where proto-institutions that pool resources can be formed by agents. The paper was presented at the Computational Social Sciences 2018 conference in Santa Fe, New Mexico in October 2018.
Katz’s research of original tax, population survey and land ownership records dating from the mid-nineteenth century showed that wealthy individuals with access to collective resources related to finance and social capital had a much higher probability of remaining in the city than the majority of citizens without such resources. If you were among the poor in Hamilton, there was a high probability that if you lived there in 1851, you would not be there in 1861.
This dynamic – that the wealthiest had dynamic stability over that decade while the poor experienced dynamic instability – suggested that some sort of institutional access (money, social ties, access to knowledge and insight, and so on) translated into stability unseen among those who did not have such membership/access.
In our model, two agents that are physically (in terms of geography) close together form a resource collector where their extra resources collected from the environment can be held. If at any time those agents are short on resources, they can draw down what is held in common by the resource collector – what we call the ‘proto-institution.’ This is the ‘spark of life’ moment for common resource use, a highly simplified form of collective function that we are exploring in order to better understand the various dynamics of socio-economic inequality. In our model, none of an agent's features predisposes it to form an alliance with another agent; the only determining factor is physical proximity with another agent, along with both agents having a surplus of wealth that they can "bank" in a proto-institution. That simple possible dynamic within the model’s environment can yield important insights.
Our early results suggest that common pooling of resources does lead to advantage for those agents that can access that resource in times of hardship, leading to an outcome where those agents that had a surplus of wealth and banked it during fair weather times were able to withdraw from that reserve during difficult times. In contrast, those agents that could not be part of a proto-institution or had exhausted their previously deposited wealth, could not weather difficult times, faced diminished odds of survival, leading to either death or migration out of their local economy.
The figure (reference figure here) provides a summary of aggregate-level emergence of inequality across two scenarios: (a) one where there are no proto-institutions (left panel), and (b) one where there are proto-institutions (right panel). From these results, we can conclude that in the scenario where proto-institutions are present: (a) the wealth inequality increases almost continually over time (note the dark line representing the median value of Gini coefficient's distribution is increasing across time); (b) the variance (the range of possible levels of inequality) tends to remain roughly the same; and, (c) extra-ordinary levels of Gini coefficients tend to appear as time passes, as seen in the appearance of outlier points towards the top of the plots in later time periods in the proto-institution scenario.
In contrast, in the no-proto-institution scenario: (a) the change in median level of wealth inequality remains roughly constant across time, after an initial few periods where it increases; (b) the variance in inequality tends to increase over time; and, (c) there do not appear to be extra-ordinary levels of inequality emerging over time.
A great deal of work has been devoted to this problem across disciplines over time. Our interest is to return to as simple and clear a moment as possible, not to provide a definitive explanation but to see how such simplifications generate insight. Our early results suggest that common pooling of resources does lead to advantage for those agents that can access that resource in times of hardship, leading to an outcome where those agents that had a surplus of wealth and banked it during fair weather times were able to withdraw from that reserve during difficult times. In contrast, those agents that could not be part of a proto-institution or had exhausted their previously deposited wealth, could not weather difficult times, faced diminished odds of survival, leading to either death or migration out of their local economy.
About
Milton Friesen directs the Social Cities program at Cardus in Hamilton, Ontario. He has taught and studied at the School of Planning, University of Waterloo and his research focuses on developing innovative ways to measure and strengthen the social fabric of neighborhoods. His recent publications include a Journal on Policy Studies and Complexity paper on computational modelling and policy design oriented to the development of more effective social infrastructure as well as a new social capital survey instrument published in The American Sociologist.
Srikanth Mudigonda directs the MS Applied Analytics (MS AA) program at Saint Louis University, in St. Louis, Missouri, USA. He teaches bachelor's and master's levels courses in computer programming, statistics, machine learning, and mentors master's students in the MS AA. His research interests include computational modeling of social and socio-technical systems, teaching and learning with technology, and related topics.