conducted. This would include a description of the sample design (who the respondents were and how they were chosen) and a copy of the complete questionnaire that gives full question wordings and the order in which they were presented (what the respondents were asked).
Elected officials are so used to employing polls in their campaigns that they are predisposed to pay attention to polls about what citizens think. They have other sources of information that provide cues about public opinion, such as letters from constituents, calls to their offices, faxes, and, increasingly, e-mail that they receive. Given that they are more knowledgeable consumers of poll information than most citizens, they are also better able to understand the sources of such information and know when they should be discounted.
Recent research shows that individuals exposed to poll results will discount them when they disagree with views they already hold, especially when the views are strongly held. This applies to elected officials as well as to regular citizens. As the Republican-controlled Congress moved to impeach President Clinton in 1998 despite repeated polls that showed public opposition to removing the president from office, Democrats cited the polls as support for a censure motion. Many Republican members, on the other hand, invoked constitutional principles and the need to fulfill their sworn obligation as a reason for ignoring these same polls. Each group was reacting to a clear message from multiple measurements on the same issue, responding in accord with their own conceptions of what needed to be done.
It is important to distinguish here between elected officials and civil servants who work in government agencies. Many government agencies are required to take public opinion into account when they embark on new projects or consider new regulations. Polls are again just one of the means that they use to ascertain what opinion is on a particular issue. For their purposes, they are often presented with polls sponsored by interest groups, so their evaluation task is often more complicated.
MARTIN, L. JOHN, ED. 1984. "Polling and the Democratic Consensus" Annals of the American Academy of Political and Social Science 472 (March). This volume is an early compendium of articles on the role of polls in the democratic process. The work is divided into four parts and consists of thirteen articles.