530 000 according to the organisers, 77 500 according to the police. As usual, nobody agrees on the number of demonstrators (against same sex marriage, for our English readers not attuned to French news) who took to the streets in Paris and Bordeaux on October 5th. And statistic has nothing to say on the subject. The French statistical office is into more noble things and NGOs are already fully occupied with Putin or Mugabe.
I will not attempt to answer that question with this blog. But the issue of the change in the number of demonstrators on this topic over a few years’ time is more interesting. We collected all polls published on the matter that you can find on the Web. Here they are (each dot in the graphic below tells the percentage of people in favour of same-sex marriage in a given poll. Two polls from Ipsos, in 2000 and 2004, are hidden by BVA polls, which results are exactly the same).
Several comments :
– In 20 of the 23 polls since 1995, French people are in favour of same-sex marriage. In most of the cases, the majority in favour is than the majority that elected president the French President in 2007 and 2012. This is not an undecided issue : there is a large majority and a minority,
– The trend is clearly upwards from 1995 to 2011, and then flat since,
– Polls in years 2012 to 2014 are rather volatile, albeit the issue does not change. Interviewees probably answer another question.…
– There are real differences across polling agencies. As usual, the answer partly depends on the way the question is asked. Ifop is asking whether “…homosexual couples […] should have the right to marry in France”, whereas CSA’s question is “Are you [in favour or not] of same-sex marriage”. Both ways of formulating the question is legitimate. Without surprise, CSA’s results are lower than Ifop’s : I am personally not in favour of smokers – at least, not of their smoke -, but I don’t believe they should be forbidden to smoke.
Where does the increase of support for same-sex marriage between 1995 and 2011 come from? Young people are much more in favour than their elders. The data I have are partial, but consistent. In 2004, an Ipsos poll giving an overall majority of 57% in favour says that 75% of 35- years old are in favour, when only 48% of the 35+ are. Let us assume that the 35- group in 2004 don’t change their mind when growing older, and the same thing for the 35s+. A quick calculation with the data from the French statistical office tells us that, in 2011, we should have 52% in favour among the 35+. A back of an envelope calculation, I agree, but which gives some direction: in 2011, similar polls gave a 59% majority for same-sex marriage among 35+ years old people, thus much more than the 52% that would come from the natural replacement of generations. In 2013, we are back to 55%, which can be compared to the 53% due to the natural replacement.
A simplified model would tell that 35- in 2004 did not change their mind when growing older, but that some 35+ did, in favour of same-sex marriage. And that some recently changed their mind when receiving their tax sheet….The percentage of French people in favour of same-sex marriage should then go on increasing, and level out in a few years around 75%.
This idea for this blog comes from articles by Nate Silver (here and here). With much more detailed data, he has built a true model on the matter. US opinion only turned in favour of same-sex marriage in 2010. But our conclusions, even based on much more fragile elements, are quite similar.
So, how many demonstrators on October 5th? Nobody will ever know. And in 5 years? The analysis of available data tells us they should be a smaller group.