Nate Silver defines herding as

the tendency of polling firms to produce results that closely match one another, especially toward the end of a campaign

On November 5 2012 Drew Linzer wrote a post "Pollsters may be herding", that stated:

[A] major concern heading into Election Day is the possibility that polling firms, out of fear of being wrong, are looking at the results of other published surveys and weighting or adjusting their own results to match. If pollsters are engaging in this sort of herding behavior – and, as a consequence, converging on the wrong estimates of public opinion – then there is danger of the polls becoming collectively biased.

I don't have a very strong reason for thinking that thread on Linzer's blog constitutes the origin of the term, but it's the earliest reference I could find. Some of the commenters on the thread asked about where this term had come from, and Linzer introduced the concept as though it was new and needed explanation.

In addition to who coined the term, I am also interested in when it was coined.

  • 1
    I asked Drew "Hello Drew, An old question at politics.stackexchange asks about the use of "Herding" as a term to describe the way in which pollsters may adjust their results to match those already published. Your post from 2012 is an early use of the word in this context. Do you think you coined the term or do you recall others using it before you? "
    – James K
    Jul 22 at 16:49

You must log in to answer this question.

Browse other questions tagged .