posted at 9:21 am on March 31, 2017 by Jazz Shaw
Last September, French physicist Serge Galam attracted some attention (and ridicule) when he applied a concept known as “sociophysics” to the American election and predicted that Donald Trump would win, despite what all the polls were saying. Feel free to wade into this quagmire at your own risk, but here’s just a taste of the explanation.
According to a model of opinion dynamics from sociophysics the machinery of Trump’s amazing success obeys well-defined counter-intuitive rules. Therefore, his success was in principle predictable from the start. The model uses local majority rule arguments and obeys a threshold dynamics. The associated tipping points are found to depend on the leading collective beliefs, cognitive biases and prejudices of the social group which undertakes the public debate.
I’m not going to pretend that I know what any of that means, but after the dust settled it turned out that Galam was correct. And now he’s sticking his neck out and making another prediction. Despite the fact that she is, if anything, even more of a divisive figure than Donald Trump, Marine Le Pen is probably in line to win the French presidency. (Daily Mail)
The scientist who predicted Donald Trump’s US election victory has claimed far-right candidate Marine Le Pen could win the French election because of voter abstention.
French physicist Serge Galam has said the National Front leader could benefit from her hard-core following who will ensure they turn out to back her.
By contrast, a substantial number of people who said they would vote for her rival may not actually go to the polls, he claimed.
Leaving sociophysics aside for the moment, Galam is describing a potential scenario which sounds surprisingly similar to what we saw on November 8th in a number of swing states, particularly in Pennsylvania. This theory seems to boil down to the idea that Le Pen’s supporters may not represent an obvious majority in raw numbers, but their enthusiasm is an indicator that virtually each and every one of them will show up on election day. By contrast, the support for her most likely opponent in the runoff election is exceedingly soft and unenthusiastic, leading to the possibility (if not likelihood) that a significant number of them will stay home.
Compare that for a moment to Pennsylvania last November, as I mentioned above. Hillary Clinton was still holding onto a slender, but clear lead outside the margins right up until the final week of the election. But those models relied on her getting out the same voter base which propelled Barack Obama to two consecutive victories in the Keystone State. That support turned out to be so soft that she underperformed in the major urban centers by a significant margin. At the same time, Trump’s voters were so fired up that he overperformed Mitt Romney in the rest of the state by an equally impressive score. It added up to enough of a difference to hand Trump the state comfortably.
Galam seems to be seeing the same phenomenon unfold in France. Le Pen’s most likely opponent in the runoff is Emmanuel Macron, but he’s not really setting the world on fire. His campaign has been dogged by scandals and he fails to excite a lot of voters because he’s widely seen as the “natural heir to President Francois Hollande.” Support for Hollande and his party has essentially collapsed and he’s seen as being “the establishment” at this point. (Is any of this starting to sound familiar?)
Could Galam be right and the national polling will turn out to produce a significant miss? Well… lightning does strike twice everyone once in a while, but we use that term because it’s still pretty unlikely. The polls are probably missing something but Le Pen is counting on an awfully big error. We can’t rule it out, but let’s just say that I’m remaining cautiously skeptical at this point. We’ll probably know more once the first round of voting is over and people are faced with the reality of a binary choice in the runoff.