The British Election Study suggests that differential turnout the most likely cause of the GE2015 polling failure
LAB supporters more likely to have overstated certainty to vote There’s an important paper just out from the British Election Study on what went wrong with the polls on May 7th. Why did it happen? The report by Jon Mellon and Chris Prosser does not give much credence to the “late swing”, “don’t knows” and “shy Tory” theories and argues that differential turnout was most to blame. Those over-stating their likelihood to vote were more likely to be those saying…