Brexit vote: absolute numbers are misleading.

As political scientists, we are not familiar with the habits of medical journals. In most analyses of voting behaviour it’s common to focus on the relative risk of voting for a party or an outcome: an analysis that focuses on absolute numbers is therefore liable to mislead. If medical journals also tend to focus on relative risks then Dorling’s analysis is also liable to mislead, notwithstanding qualifications in relation to “differential turnout and the size of the denominator population.”