It is striking: Numerous elections and referenda in the last years have ended with extremely close outcomes, such as the American presidency election in 2016 or Brexit, which was in the same year. At the same time, there is an increase in the polarization of the political parties – standpoints are becoming more radical, the tone is becoming sharper, and the discussions are more emotional. Is there a connection between these two developments? “According to our findings, yes,” says Professor Stefan Borndholdt from the Institute for Theoretical Physics at the University of Bremen. “The most important findings from our research: An increase in the stating of political messages that could be potentially repelling results in a higher probability that there will be a close outcome when the decision is made.” The Bremen physicists have now published their results in the biggest and oldest physics journal; Physical Review from the American Physical Society.
Explaining Human Behavior Using Sociophysics
Borndholdt and his colleagues carry out research in the field of sociophysics, alongside other areas. Said field describes and explains the rules of behavior of large crowds using methods of physics. In the case of the close election outcomes, the Bremen researchers applied mathematical probability statistics: the so-called Voter Model. “This ‘Voter Model’ is very similar to the physical model of the alignment of magnetic atoms,” explain the university lecturer. “We used it in order to investigate the noticeably close outcomes of political debates and to find possible reasons for this.”
Voter Model Forms the Basis
The Voter Model assumes that there are only two opinions for each situation – for example during the election campaign of a political party: Agreement or disagreement. “In this model, two sides meet. Which of the two opinions is in fact enforced, is decided by tossing a coin.” This mean that it is chance if person A can be convinced by person B or if they stick to their original opinion,” says Borndholdt. “The interesting thing is that if you repeat the meeting of the sides and the chance decision-making long enough, the opinion disappears after a while and the other opinion wins.” The Voter Model is an extremely simple yet robust model. “One of the two opinions is always enforced yet it is chance which one of them it is.” This works in the same way in the fashion sector, for example, and provides indications of why different colors, such as green in one season and ash gray in the next, catch on in the “fight of the colors”.
Brexit Polls from Seven Years Investigated
“We have taken a look at polls from seven years regarding the Brexit decision on June 23, 2016. Whilst the numbers of people for and against leaving the EU lay far apart – initially with the majority being for leaving – after two years of the opinion-forming process, the majority ratios came continually closer.” This surprises Bornholdt and his colleagues: According to the Voter Model, one of the two opinions should have caught on in the long-term. “In reality, it would not have been 100:0 as in the model but rather 70:30 or similar.”
Noticeable with regard to the Brexit polls: Alongside the circumstance that both political sides were usually around the same size, there was an unusually high potential of undecided people. “We have now incorporated this situation into the model. We have added the middle status “undecided” and a further two possibilities for changing one’s mind. One the one hand, people who had already decided now have the chance to doubt their assumed opinion and go back to the undecided camp. On the other hand, we met the needs of the observation that the discussion continually became more intense,” explains the physicist. Passionate debates, demonization of counter opinions, personal attacks – all of this was observed in Great Britain. The same can be said for the American presidency election.
Repelling Messages Integrated into Model
The circumstance that changed opinions can also have a repelling effect and can almost push a communication partner into the opposite camp rather than win them over for one’s own beliefs was also incorporated into the model. “If we now continually increase the number of the exchanged opinions that have a repelling effect in the model, there is suddenly a critical value. If 25 percent or more of the exchanged opinions have a repelling effect, there is no longer a winner.” Thus, if the political messages are so provoking that they actually strengthen the opposition in more than one in four cases, a close outcome of the decision is extremely likely, according to the Bremen researcher’s calculations.
“The Voter Model then undergoes a phase transition and behaves entirely differently,” says Stefan Borndholdt. “If one runs the model under these conditions for long enough, a result close to 50:50 occurs.” The conclusion: The more hate speech that is used in a debate over a long period, the more likely it is that there is a closer polarization when both groups have roughly the same size. “This effect, which we studied using a simple computer model, can in our opinion also occur in real life and may be an explanation of the close election and referenda outcomes.”
Models Shown Possible Effects
With their models, the experts from the Institute for Theoretical Physics have repeatedly shown that the regulations in day-to-day life, which are created by people, develop more effects than originally planned. Bornholdt’s research group at the Institute for Theoretical Physics has found out that the prescribed installation of “intelligent electricity meters” in new and completely renovated buildings, which has been in place since 2010, possibly does not result in saving electricity. Due to the demand for electricity at particular times, it is possible that bubbles will form in the electricity market – much like in the stock exchange, which could lead to an overload of the network and “blackouts”. In a different study, the scientists proved that the repeated changing of gas prices within a day generally leads to a higher price level.
Sebastian M. Krause, Fritz Weyhausen-Brinkmann, Stefan Bornholdt: „Repulsion in controversial debate drives public opinion into fifty-fifty stalemate, Physical Review E 100 (2019) 042307. https://journals.aps.org/pre/abstract/10.1103/PhysRevE.100.042307
Prof. Dr. Stefan Bornholdt
Institute for Theoretical Physics
University of Bremen
Phone: +49 421 218-62060
Email: bornholdtprotect me ?!itp.uni-bremenprotect me ?!.de