The People Have (been) Spoken: Exit Polls and Public Agendas
Today, the day after the midterm elections, we are swimming in exit poll data. How that data was produced and what it actually means is not as obvious as some commentators would often have us believe when they tell us the people have spoken.
When we are told by CNN that 42% of voters polled cite corruption as a reason for voting the way they did, 40% terrorism, 39% Economy, and 37%
It’s more interesting to media and politics scholars when the questions are open (“So what issues are on your mind?”). Then they can compare those responses to the media agenda itself. The media agenda is what stories the news actually covers out of the millions of things going on in the world every day. If there is a strong correspondence between the responses about important issues and the findings of what appears to be the media agenda, then we see that the media have very likely played an important agenda-setting role for “the public”; they have primed the audiences about what to consider as important. Even there, sometimes studies may find that respondents are influenced not just by the news media but also or to varying degrees by opinion leaders they know, people who will influence them by one degree of separation from the news. These predictions are quite difficult given the wide variety of media diets people have these days. But some studies try to take that into consideration.
For example in a 2004 study, the Pew Center on the Public and the Press found that roughly 21% of young people ages 18-29 received their political campaign information from late night comedy shows such as “David Letterman, Jay Leno, Saturday Night Live or The Daily Show.” There could be other TV shows that would diversify and break down the portrait of media consumption further, but these are the ones they ask about and they do tell us that people watch them. But we don’t know how many other shows they get information from in addition. Incidentally, this fascinating study maps out internet, TV, cable, radio news consumption about the 2004 campaign.
What is more, these representations/assertations of public opinion are then sometimes found to have secondary effects on those who consume the report of public opinion, one of the most common being the bandwagon opinion effect, where a person takes up a position because he/she doesn’t want to be associated with something perceived as unpopular.
Returning to the exit polls for the recent midterm election, we can see that the knowledge they produce is as good as the questions they ask. At a deeper level of analysis, one can even wonder if we can ever have completely reliable information about human motives for voting, among other actions. Is it always obvious why people do things? Do they rationalize in retrospect? Besides creating a position, an opinion that may not have existed in any strong form or even at all before the question, polls also transform a lot of singular responses into one big number which becomes “the public” in public opinion. However the term is a bit misleading, since when people actually do get together and discuss what they thought about issues, issues which bring them together as a public, the answers might very well be different. Groups work according to different dynamics, especially when they follow institutional rules for debate and considering multiple points of view before coming to a position or opinion. Finally, as with most forms of question and answer social science, polling is subject to the strange behavior that people manifest when they are asked questions about why they care about or do particular things.
Exit polls show responses to questions; they do not necessarily reflect accurate public opinion of what is most important in an open-ended way. When politicians read polls and then declare, “The people have spoken,” what publics have said exactly for policy-making is not nearly as obvious as one might assume.