First Brexit, Then Trump. Is Le Pen the Next Political Earthquake?
In today’s political climate, “business as usual” no longer applies. Disruption is in, the status quo is out. Once-reliable voting blocs are in flux. And, when it comes to predictions, topline polling can’t always be taken at face value. This certainly proved true in the election of Donald Trump and the passage of Brexit. And now, the situation in France seems even more unpredictable, after months of scandals, late entrants, and a last-minute terrorist attack.
Many of the factors that influenced the U.K. and U.S. elections — populist anger, anti-immigrant and nationalist rhetoric, a lack of enthusiasm for establishment candidates — are shaping the presidential campaign of far-right candidate Marine Le Pen in France. But despite the similarities, most pundits and polls predict she is unlikely to win.
We think that’s naive. Applying our experience with data science and polling, we endeavored to use the lessons learned from the U.S. election and Brexit to take a deeper look at the question currently on the minds of those following the French election:
Can Le Pen win and, if so, how?
While the majority of simulations we ran indicated that a Le Pen run-off victory is still a long shot, we did identify some realistic scenarios that could tip the election her way. Here are some key factors to watch:
Undecideds. France’s first (of two) rounds of presidential voting is around the corner, and according to opinion polls, as much as a third of voters are still uncertain of their choice. To date, Le Pen has polled consistently well while establishment candidates have struggled: incumbent Hollande chose not to run for re-election, a former president lost in the primary, and the mainstream conservative candidate Francois Fillon has become engulfed in scandal. Fillon’s lead has recently evaporated, giving new momentum to an untested centrist candidate, Emmanuel Macron. And in the past couple of weeks yet another atypical candidate, Jean-Luc Melenchon (far left), has been surging in the polls. In this constantly evolving atmosphere, French voters are taking a wait-and-see approach, potentially leaving an opening for the far right.
In the U.S., undecideds broke late and overwhelmingly for Trump in many key states, including by a 2-to-1 margin in Wisconsin and a 3-to-2 margin in Florida and Pennsylvania. If late deciders move towards Le Pen in a similar fashion, this could give her a winning edge in the run-off, especially if she faces Fillon.
- Le Pen vs. Fillon: Fillon’s late-breaking nepotism scandal is dominating the news at a crucial point in his campaign, similar to the Clinton e-mail story. This could potentially dampen turnout with his base, as well as with left-leaning voters who did not support him in the first round but would likely never vote for Le Pen. Conversely, Le Pen seems immune to similar charges with her followers, just like Trump. Since the beginning of the year, Le Pen has been polling at a remarkably consistent 26 percent plus or minus a single percentage point, giving her the lead in a crowded field despite several unfavorable stories and investigations. In contrast, Fillon dropped 5 percentage points in a single day once news of his scandal broke and has not recovered. In one plausible Le Pen win scenario, lower turnout for the groups least favorable to Le Pen (we estimated 15 percent), combined with undecided voters breaking to Le Pen 20 percent more than currently estimated is enough to put her over the top.
- Le Pen vs. Macron: If Macron ends up in the run-off, polling predicts he would do better. Nevertheless, it is important to keep in mind that he is not as well known, and his support is more tepid. Eight out of 10 (80%) Le Pen voters are certain of their position, while over half of Macron voters are uncertain. Intensity can have a huge impact on turnout.
Working class voters. In the U.S. Presidential election, turnout was much higher than forecasted among social and demographic groups favorable to Trump and lower than predicted for groups favorable to Clinton, based on polling and 2012 outcomes. For example, Trump ultimately won white working class voters by a 39-point margin (67% to 28%), performing much better than pre-election polling had predicted and doing a net 14 points better than Mitt Romney did in 2012. By simulating similar changes in turnout and support for Le Pen, we were able to find several scenarios in which she would win, including one based on working class support — a group she currently polls strongly with. If there’s a 10 percent increase in support among voters without a Bachelor’s degree for Le Pen (in line with how Trump did compared to pre-election polling) and a 10 percent decrease in turnout from voters with a higher education, she can garner 50.2 percent of the vote in the run-off and become the next president of France.
To be clear, we are not predicting a Le Pen victory. The data indicate she would lose in a run-off in most scenarios. However, it is essential to view French pre-election polls in the context of a political season marked by disruption.
As a result of our analysis, we believe France and the rest of the world should not assume and act as if Le Pen cannot win. In fact, there are striking similarities between Le Pen and Trump’s trajectories, and windows of opportunity that can lead to yet another political upset that further empowers the far right. This time, we hope our findings serve as a wake-up call to forces who oppose the far right to get their act together to prevent this now feasible outcome.
Ben Tulchin is the President of Tulchin Research and served as the pollster for Bernie Sanders’s Presidential campaign
Marion Azoulai is a Senior Data Scientist at Alpine Data
