1

A France24 article from June says

RN’s performance at 90 seats is a “seismic event; an extraordinary result for them”, said Paul Smith, a professor of French politics at Nottingham University. “There were no polls predicting this and I haven’t seen anybody predicting it. Le Pen was looking washed up after the presidential second round; so many people thought that was it for her – and she herself wasn’t really campaigning for the législatives. But, clearly, that wasn’t it.”

The true number was 89 seats for RN, but still that was indeed far above what polsters predicted; in the polls from June it seems the median upper limit predicted was 45 or 50 (eyeballing it in that wiki table), with just one pollster predicting 60 seats at the most for RN.

So, why did pollsters do so poorly in predicting RN's seats?

the gods from engineering
  • 158,594
  • 27
  • 390
  • 806
  • Vaguely related, I suppose https://politics.stackexchange.com/questions/13075/why-were-pre-election-polls-and-forecast-models-inconsistent-with-the-election-o – the gods from engineering Sep 28 '22 at 20:23
  • looking at the highes-voted answer to the linked question, it seems to be also applicable here; it presents multiple non-US examples of the same trend (underforecasting for conservative parties). – Danila Smirnov Sep 29 '22 at 03:16
  • If someone had the actual answer, they could get a nice gig working at a polling company. And since the polling companies are still getting it wrong, I don't know if that's all that answerable here. – Italian Philosophers 4 Monica Sep 29 '22 at 20:28

0 Answers0