Political polls are surveys conducted in order to predict the results of an upcoming election. In the UK, they usually involve 1000-2000 participants, although this number is sometimes higher. The people being surveyed are asked who they are planning to vote for, or, in referenda, what option they are voting for, as well as how likely it is that they will vote. Their gender, age and place of residence, amongst other factors, are then taken into account so that the company conducting the polling can ensure that the poll is representative of the population as a whole. In Britain, there is an array of polling firms, from well-known ones like YouGov and Ipsos MORI to lesser known companies such as Panelbase and TNS. As well as asking questions about peoples’ voting intentions, these companies also ask other political questions, such as who the respondents think would make a better prime minister. Occasionally, they pose questions completely unrelated – polls on issues ranging from religion to Eurovision have also been produced in the last year.
However, the reputation of the British polling industry has suffered greatly in recent years. There were 6 separate polls conducted partially or entirely on 22 June, the day before the EU referendum took place. 4 of these polls predicted the Remain side winning, while only 2 of the polls predicted, accurately, that the Leave side would win.
Similarly, on 8 November, every single aggregate poll tracker was predicting that Hillary Clinton would win the popular vote by margins ranging from 1.9% to 5.3% and that she’d also win the Electoral College vote – but only the first prediction was right.