Your source for the latest research news

Sunday, 2 February 2020

The laws of physics that explain political polarization in elections

The study of political polarization, which has emerged around the world, may have much to gain from the use of some tools and formulas used by physics. [Image: MIT]

It may seem surprising, but theories and formulas derived from physics can be useful tools for understanding how democratic elections work, including how these systems fail to deliver on their promises and how they can be improved.

Alexander Siegenfeld (MIT) and Yaneer Bar-Yam (New England Institute of Complex Systems) took political-electoral data and analyzed it using various well-known laws of physics as tools. And they demonstrated how these laws can be used to describe the behavior of the data.



The application of several of the physics formulas to the US electoral system revealed that the elections went through a transition in 1970, from a condition in which the election results reasonably captured the electorate's greatest political preferences, to a period of increasing instability, in which very small changes in voter preferences have led to significant changes towards more extreme political results in both directions.

The two physicists found that the Ising model , developed to explain the behavior of ferromagnets and other physical systems, is mathematically equivalent to certain election models and accurately describes the onset of instability in electoral systems.



In this regime of "unstable" elections, "a small change in voter opinion can dramatically alter the outcome of the election, just as the direction of a small push on a rock at the top of a hill can dramatically change its final location," said Siegenfeld .

"What happened in 1970 is a phase transition just like boiling water. The elections went from stable to unstable," added Bar-Yam.

Negative representation

The analysis shows that this instability can be associated with an unexpected situation in which the results oscillate in the opposite direction of how people's real preferences are changing. In other words, a small movement in the predominant opinions towards the left can result in a result more to the right and vice versa - a situation that the researchers call "negative representation".

"Our country seems more divided than ever, with the election results looking like a pendulum swinging with increasing strength," said Siegenfeld.

This long-term shift from a stable electoral situation to one marked by instability is similar to what happens with ferromagnetic metal exposed to a magnetic field, adds Siegenfeld, and can be described by the same mathematical formulas.

Predict the whole without knowing the parts

But why can the derived formulas for such different subjects be relevant to the political field?

Siegenfeld says that it is because in Physics it is not always necessary to know the details of the underlying objects or mechanisms in order to produce useful and significant results. He compares this to how physicists were able to describe the behavior of sound waves - which are essentially the aggregate movements of atoms - with great precision, long before they knew about the existence of atoms.

"When we apply physics to understand the fundamental particles of our Universe, we don't really know the underlying details of the theories," he said. "However, we can still make incredibly accurate predictions."



Likewise, researchers do not need to understand the reasons and opinions of each individual voter in order to conduct a meaningful analysis of their collective behavior.

As the pair's article states, "understanding the collective behavior of social systems can benefit from methods and concepts in physics, not because humans are similar to electrons, but because certain behaviors on a large scale can be understood without understanding small-scale details."


Bibliography:

Article: Negative representation and instability in democratic elections

Authors: Alexander F. Siegenfeld, Yaneer Bar-Yam

Magazine: Nature Physics

DOI: 10.1038 / s41567-019-0739-6

No comments:

Post a comment