In three of four surveys Pew has conducted since spring, estimates from landline samples alone produced “slightly more support for Republican candidates and less support for Democratic candidates, resulting in differences of four to six points in the margin”. One survey showed no difference between the landline-only and combined samples.

Pew says its data shows that non-coverage bias in telephone research “is as large, and potentially even larger, than it was” in 2008 pre-election polls.

A quarter of US households are now thought to be cellphone-only and cell-only adults have been found to be “demographically and politically different from those who live in landline households”, says Pew – as a result of which election polls that rely only on landline samples “may be biased”.