| NAVIGATION | |||
| Reading | Test Material | Discussion | Links |
O'Connor and Sabato, Chapter 11, (pages 385-416)
These are clues to the multiple choice questions on the final from this chapter:
Polling and Predictions: Does my vote really count?
One of the enduring controversies of modern politics is about the effect of network polling announcements and their effect on "western" returns. Or putting it another way, if candidate X has been declared the winner by CBS, NBC, ABC, FOX, etc, why should supporters of Candidate Y vote (or do they just stay home)? Well the last time I checked, the President was elected by a vote of the electoral college after a vote by the public to determine electors, not by the pundits of the networks. So how do the networks make their predictions?
The question above leads me to comment on the types of polls. Polling companies that want to be accurate, use a technique well described in your text called "quota sampling". This method insures that the sample will "look like" the group to be sampled. Generally, the theoretical accuracy of these polls is plus or minus 3 points (a 6 point spread). However, their typical accuracy is much greater than that. Over 30-40 years, most of the major poll have been within 1% of the final outcome. In fact, with the networks making one of the biggest errors ever, the traditional pollsters had it just about right, namely Al Gore by about half a percent. The final figure was pretty close to that.(Remember, he did win the popular vote.)
The networks use a different type of poll called an exit poll. It is a type of poll where the questions are asked of voters exiting the ballot place. Networks usually don't just rely on exit polls alone, but when they are consistent with other polling techniques, exit polls are generally reliable. So, what happened!?? First, the election (particularly in Florida) was as close as close can be. Even a small margin for error won't be good enough when the margin between candidates is .01%. Nobody is that good. Secondly, apparently a significant number of voters miscast their votes in some way (mismarked, double marked, etc., you know....the "chad" thing). They would have come out of the polls and said they voted for Gore and not know they had messed up their ballot. This would have unintentionally misled the pollsters. Third, there were machine error problems. These are usually assumed to be neutral, but perhaps were not this time. The networks are currently seriously re-examining their polling techniques especially the less scientific "exit polls" to see what changes need to be made. They will probably make some serious corrections and get a better read on the voters. Also, they will learn to be a bit more careful in elections that are this close.
But if they are going to continue to make predictions (and they are), what effect on voting will it have. Surprisingly little it seems. The few studies that have been done on the effect of early predictions of western voting have shown little if any NET effect. The reason I emphasize "net" effect is that it is possible to find folks who said to themselves, "What the hell, its over so my vote won't matter anyway!" One has the sneaking suspicion that they were not enthusiastic voters to begin with. In addition, they might well be balanced by those who have become motivated by an adverse prediction ("those s.o.b.'s can tell me, etc). We always had before us the examples of Hawaii and Alaska who always finish voting long after the outcome is settled no matter how close the election. There, turnout has seemed to be as good (or as bad) as say, North Carolina or Pennsylvania. According to the critics, they should have forsaken the election, but they didn't. Why not?
Voting is an habitual activity, that is, once you start you tend to keep voting even in elections that are not that close. The reverse is also true. Folks who wish not to vote can find any number of excuses such as the lame "I don't want to be called for jury duty". But there is another more devious problem, not with polling per se, but with the use of polling.
Modern politics has seen the willingness of Presidents to rather slavishly follow the polls. It has also empowered members of Congress to oppose good policy just because it is currently unpopular. This is reflected in the criticism that most "leadership" today is "short term" thinking, paying too much attention to the polls. Lincoln and Truman were never wildly popular during their great presidencies. Nevertheless, they sometime pursued unpopular policies which they (and history) knew to be right. Bill Clinton did something similar with NAFTA. We'll have to wait to see if the president was wiser than the public. Governing officials need to be able to make unpopular decisions and then have a decent period of time to develop support for wise but unpopular policies. I dare say, that we might still be a segregated country if we had just relied on the "popularity" of integration policies. I know candidates need to win elections, so there need to be a balance between what needs to be done and what is popular in the polls.
For more information on the effect of polling try:
Asher, Polling and the Public: What Every Citizen Should Know, 4th ed. CQ Press
This area contains a few links to sites that have material that applies to the subject. I hope to expand this as we go along and invite you to send links to me when you find interesting material that you think is relevant. Good Hunting. The text site is always included as the first link.
| LINK SUBJECT |
|---|
| The text site |
| The Gallup Poll |
| The Roper Poll |
| The Pew survey |