|
|
|
|
|
|
|
Click here to see our elections web page
*Why it's not a free country anyway
*Examples where the united $tates has not respected majority rule before
*See our article refuting a certain line of approach to seeing Florida as a fraud
*See our article on Zogby as a case of how capitalism influences polling
by a contributor
Currently on the Internet there is a dispute concerning the 2004 elections with Republican activists saying the exit polls are biased toward Democrats. Some have even suggested that rigging exit polls was part of a Democratic plot to influence voters who had not voted yet.
By leaking incorrect results to the media during the day of the election it may be possible to influence votes. The fact that such a conspiracy is thinkable is a poor reflection on U.$. voters and the media vultures looking for the latest news to make a buck off. If the Democratic Party did carry out an exit poll rigging, then that would also reflect poorly on the regard of Democrats for civic duties in Amerika. While we at MIM counsel the public not to vote and spend the time in political activism instead, we oppose the manipulation of exit polls as well. As a result, we are calling for the release of exit poll data including "proprietary" information on sample design and weighting schemes and any documentation of the polls that existed prior to the poll as well as any post-hoc explanations of what may have happened to cause Kerry to win the exit polls and not the vote.
Polls opened November 2nd at 6:30 AM in Ohio, 7 AM in Florida, and 7 AM in Pennsylvania. Exit poll data for a past tax levy in Cleveland indicate that within a given locality, time of voting is related to demographics.(1) At different times of the day, even well-designed exit polls reflect particular demographics. Busch and Lieske suggest that time of voting is "moderately correlated" with employment status. "Employed people, with fixed hours of work, are more likely to vote in the early morning and late afternoon. By contrast, unemployed voters, a category that includes housewives, students, and retirees, have a more flexible day and thus vote more often in the midmorning and noon period" (p. 99).
Ohio had an above-average unemployment rate for September, but still voted for Bu$h. However, even in Ohio, early exit poll reports showing a KKKerry lead could reflect a great proportion of unemployed people voting--the assumption here being that unemployed people tend to vote against the incumbent President. Several people are wondering why Kerry's lead in the exit polls narrowed as the day went on. This may partly explain why. Unfortunately, exit pollsters did not ask about employment status.
Although final exit poll results have been extremely reliable(3), early exit poll reports may influence vote behavior. "The only time the exit polls can have an effect is when they change voters' perceptions about the closeness of the race"(2).
So, conspiracy theorists are saying that the exit polls were rigged to show a Kerry blow-out. It is no surprise that candidates can possibly benefit from exit poll manipulation.(4) And clearly, early exit poll reports had an impact on people's perceptions(5), if not their behavior, even going as far as affecting the stock market(6). The question is, how could early exit poll reports showing a Kerry blow-out have helped Kerry assuming that Bu$h really did have a lead? If Bush really didn't have a lead? Big lead? Small lead?
Contrary to the belief that "bandwagon effect" assumption that people will vote for the candidate who seems to be winning by a substantial margin, "although some studies found evidence that voter preference shirts toward the candidate leading in the polls, other studies found evidence that voter preference shifts toward the candidate trailing in the polls"(7). It seems that "polls may cause some voters to switch their vote in one direction and others to switch in a different direction"(8). Seemingly confirming Sudman(2), Morwitz and Pluzinski mention research indicating that reports of early election results decrease voter turn-out. The question is, which candidate is disproportionately affected.
The decreased turn-out issue aside, whether there is a bandwagon effect or an underdog effect for people who do go to vote depends on the characteristics of the individual voters. Early exit poll reports showing a Kerry blow-out could have actually discouraged Kerry supporters from voting. To encourage Kerry-supporters to vote, it might have been in Kerry supporters' interests to rig the exit polls to show a slight Bush lead. Apparently, this didn't happen.
"Previous studies have found that voters who favor a given candidate are more likely to expect that candidate to win the election"(9). Morwitz and Pluzinski do not directly deal with individuals who end up not voting, but we can make these hypotheses:
1. People who intend to vote for Kerry, expect Kerry to win, and see an apparent Kerry blow-out, will not bother going to vote.
2. People who intend to vote for Kerry, but do not expect Kerry to win, and see an apparent Kerry blow-out, will not bother going to vote.
3. (Based on Morwitz and Pluzinski's cognitive dissonance hypothesis #1) People who do not intend to vote for Kerry, but expect Kerry to win, and see an apparent Kerry blow-out, may vote for Kerry.
4. (Based on Morwitz and Pluzinski's cognitive disonnance hypothesis #3) People who do not intend to vote for Kerry, and do not expect Kerry to win, and see an apparent Kerry blow-out, will disregard the early exit poll reports. They will still vote for Bush.
From the viewpoint of Kerry-supporters who are planning to rig the exit polls, the third effect must be weighed against the first and second, and possibly the fourth if the apparent Kerry blow-out mobilizes Bush-supporters to vote for Bush. There is no obvious reason for the Kerry-supporters to rig the exit polls to be in "favor" of Kerry unless Kerry already had a substantial lead by accident, in which case simply discouraging voter turn-out by portraying a Kerry blow-out might be to Kerry's benefit. I have not seen any evidence that portraying a Kerry blow-out would have helped Kerry if Bush was actually in the lead by a substantial margin.
It would have been better to portray a slight Bush lead, or even a slight Kerry lead, to encourage Kerry-supporters to vote for Kerry. For Kerry-supporters to rig the exit polls to portray a Kerry blow-out would have been a gamble if, in reality, Kerry had only a small lead, Bush had a small lead, or Bush had a big lead.
If, nation-wide, Bush actually had a big lead during the day (before the polls closed), it is not clear how Kerry-supporters' portraying a Kerry blow-out would have helped Kerry. However, a situation in which it would make sense for Kerry-supporters to portray a Kerry blow-out is where they expect the population (prior to seeing early exit poll reports) to intend to re-elect Bush, in which case they have nothing to lose by rigging the exit polls to show a Kerry blow-out. In fact, academics predicted, before election day, that Bush would win ("Who Correctly Predicted the Outcome of the Election?: A Report Card," http://hnn.us/articles/8339.html). It would make the tallies look suspicious and cast doubt on a Bush victory and at the same time discourage some Bush votes and encourage some Kerry votes.
Localized misreporting of exit polls, to portray a Kerry blow-out probably would have had a bigger impact as a form of voter intimidation than actually rigging exit polls nation-wide. Picture this: Kerry-supporters expect that an overwhelming majority of people in a precint who intend to vote, intend to vote for Bush and expect him to win, so the Kerry-supporters misreport exit polls to people in the precint to portray a huge Kerry blow-out and discourage these people from voting. This discourages more Bush-supporters from voting than Kerry-supporters.
Notes:
1. Ronald J. Busch and Joel A. Lieske, 1985, Does time of voting affect
exit poll results?, Public Opinion Quarterly, 49(1), 94-104
2. Seymour Sudman, 1986, Do exit polls influence voting behavior?, Public Opinion Quarterly, 50(3), 331-339
3. Warren J. Mitofsky, 1998, Review: was 1996 a worse year for polls than 1948?, Public Opinion Quarterly, 230-249
"Since the networks formed their exit poll pool in 1990 they have covered about 500 races. Over half were projected from exit poll results. The only incorrect projection by the pool was in the New Hampshire Senate race in 1996. This lone error was corrected by the networks on-air two and a half hours after the mistake was made" (p. 247).
4. Trilochan Sastry, 2004 April 28, "Exit polls can unfairly affect
outcomes," Economic Times,
http://www1.economictimes.indiatimes.com/articleshow/643786.cms
5. David Bauder, "News organizations to review reliability of exit-poll
data," Associated Press, Tennessean,
http://www.tennessean.com/elections/2004/archives/04/11/60914538.shtml
6. Cynthia L. Webb, 2004 November 3, "Bloggers let the poll cat out of
the bag," washingtonpost.com,
http://www.washingtonpost.com/wp-dyn/articles/A21932-2004Nov3.html
7. Vicki G. Morwitz and Carol Pluzinski, 1996, "Do polls reflect opinions or do opinions reflect polls? The impact of political polling on voters' expectations, preferences, and behavior," Journal of Consumer Research, 23(1), 53-67, pp. 53-54
8. Ibid., p. 54
9. Ibid., p. 59