This starts to explain the modal divide
After all the discussion during the week about why the phone and online polls are showing such different pictures I’ve been examining the detailed data from he last eight polls. The area where the two modes most divide is with the oldies – the group that, as we all know is most likely to turnout on June 23rd.
The chart above shows the turnout weighted percentage of those oldies expressing a voting intention in the last eight polls – four internet and four online
The very striking thing about the chart above is that the “oldies mostly going for LEAVE” narrative is driven by the internet surveys which, of course, have been showing very different EURef figures. . Just look at the big difference between their LEAVE numbers for the oldies and the phone firms.
Historically the internet pollsters have struggled with samples in the over-65 segment and those that do take part are clearly tech-savvy and might not be as representative of the full age group.
With the phone pollsters its a different story. They have far fewer problems filling the 65+ segment of their samples and, invariably, they are over-represented with their views have to be scaled down. Oldies are much more likely to be at home ready to receive phone calls.