2008 and 2004 Presidential Exit Poll Discrepancies Compared

Introductory Summation:

In 2008 the exit poll discrepancy was considerably smaller than in 2004, but it was still well outside the margin of error. I won’t calculate an exact number, since we don’t have all the data yet. But it’s safe to say that the difference is very unlikely to be explained by chance alone.

The fact that pre-election polls [1] provided an estimate very similar to the exit polls in 2008 (The Obama lead was a little bit less in the pre-election polls, but it was surging upwards in the last couple of days, so probably the two were about equivalent) makes it even more likely that they were both accurate.

So that leaves two possibilities: Exit poll bias (and pre-election poll bias as well) or impaired election integrity – that is, election fraud.

Consequently, EDA undertook an effort yesterday to capture exit poll statistics from all major statewide races (President, Senator, and Governor) prior to “adjustment” of the statistics to match the official election results. (Once the statistics are “adjusted” to match the official election results they are worthless for the purpose of assessing the exit poll discrepancy because the “adjustment” erases the discrepancy.)

"What all this means is that, as in 2004, the Democratic candidate performed much better in exit polls than in the official vote count, and the difference was especially large in critical swing states."

To read the complete article, click here [2]

Source: Time for Change's Journal, posted 11.05.08 06:22 PM
http://journals.democraticunderground.com/Time_for_change/383 [3]

Comparison of the 2008 and 2004 Presidential Exit Poll Discrepancies and their Significance

By Dale Tavris, EDA Coordinator for Election Data Analysis

The presidential exit poll discrepancy of 2004 generated a great deal of controversy in some quarters because it raised the suspicion of a stolen presidential election. There were two types of extreme and opposite reactions to that controversy that I believe were unhealthy to our democracy.

At one extreme were those who chose to ignore it. These were people who, if they were aware of it at all, simply felt that it wasn’t important. Probably some of these people just couldn’t bring themselves to acknowledge even the possibility that a presidential election could be stolen in our country. Or maybe many of them believed that it is “bad sportsmanship” to question the results of an election once it is officially declared final.

At the other extreme were those who felt that the 2004 exit poll deficiency was proof of unlimited power by the Republican Party to steal elections. Consequently, they believed, any traditional efforts on behalf of Democratic or other liberal political candidates is meaningless in the absence of conversion to an election system that involves nothing but hand counted paper ballots.

The reality of the situation is somewhere in between those two extremes. Large exit poll discrepancies should be taken very seriously and investigated to ascertain their cause – especially with an eye to evaluating the possibility of wide scale election fraud. But at the same time it is highly counterproductive to assume an unlimited capability of Republicans for election fraud and to abandon traditional efforts to win elections.

In this post I will briefly describe the presidential exit poll discrepancy of 2004, compare it with what we know – so far – of the presidential exit poll discrepancy of 2008, and discuss the possible reasons for and importance of gaining a better understanding of the causes of these exit poll discrepancies.

A brief description of the 2004 presidential exit poll discrepancy

The 2004 Edison-Mitofsky national exits polls predicted very different results than the official Presidential election results. Whereas Bush won the official results by 2.5%, the exit polls predicted a Kerry victory nationally by 3% – a 5.5% difference.

In addition, state exit polls predicted a Kerry victory in four states that Bush won – Ohio, Iowa, New Mexico, and Nevada – and a virtually even race in Florida, which Kerry lost officially by 5%. Of these states, the difference between the exit polls and the official results (which we refer to as “exit poll discrepancy” or “red shift”) were statistically significant (beyond the margin of error) only in Ohio and Florida.

In Ohio, Kerry lost officially by 2.5%, while winning the exit poll by 4.2% a difference of 6.7%. Winning either Ohio or Florida would have meant an electoral victory for Kerry. On the other hand, none of the states that Kerry carried in 2004 were predicted in the exit polls for Bush. None of this is controversial or denied by any knowledgeable person.

What we know so far about the 2008 presidential exit poll discrepancy

The Election Defense Alliance [4] (EDA), for which I work as a volunteer [5], was established shortly after the 2004 Presidential election, partly (or solely) in response to what its founders perceived as a stolen Presidential election. Their perception of a stolen election was based largely (or solely) on the presidential exit poll discrepancy.

Consequently, EDA undertook an effort yesterday to capture exit poll statistics from all major statewide races (President, Senator, and Governor) prior to “adjustment” of the statistics to match the official election results (Once the statistics are “adjusted” to match the official election results they are worthless for the purpose of assessing the exit poll discrepancy because the “adjustment” erases the discrepancy.)

I was assigned three states to monitor and document – Pennsylvania, Georgia, and North Carolina. In addition, I periodically peeked at some other results of major interest. When very early in the evening I noted large Obama exit poll leads in Virginia and Ohio, a medium sized lead in Indiana and a smaller lead in North Carolina, I was elated, believing that all this signified an almost certain Obama victory – so I posted my opinion [6] on this.

The exit poll information I had at the time was approximately the following:

Indiana: Obama by 5
Virginia: Obama by 9
Ohio: Obama by 8
North Carolina: Obama by 3
Pennsylvania: Obama by 15
Georgia: McCain by 2
National: Obama by 8.3

As it turned out later (which is not surprising), as in 2004, the Democratic Presidential candidate performed substantially better in exit polls than in the official vote count:

Official vote margin and discrepancy between exit poll and official vote count

Indiana: Obama by 1; exit poll discrepancy of 4
Virginia: Obama by 5; exit poll discrepancy of 4
Ohio: Obama by 4; exit poll discrepancy of 4
North Carolina: Obama by 0; exit poll discrepancy of 3
Pennsylvania: Obama by 11; exit poll discrepancy of 4
Georgia: McCain by 5; exit poll discrepancy of 3
National: Obama by 6.1; exit poll discrepancy of 2.2

Most of these exit poll discrepancies are beyond the margin of error or very close to the margin of error. That national exit poll discrepancy, though smaller than the others, is well beyond the margin of error because the sample size is much larger than for the state polls.

What all this means is that, as in 2004, the Democratic candidate performed much better in exit polls than in the official vote count, and the difference was especially large in critical swing states.

Possible reasons for discrepancies between exit polls and official vote counts

When exit polls differ substantially from official election results, there can be only three reasons (or combination thereof):

1. Random error, or chance
2. Biased polls
3. Impaired election integrity

The first step in the assessment of any statistical discrepancy is to assess the role of chance in producing the discrepancy. The likelihood of the discrepancy between the 2004 national exit polls and the 2004 official national results occurring by chance was estimated by Jonathan Simon and Ron Baiman [7] as being close to one in a million.

The original response to the Edison-Miftofsky report [8] by US Count Votes (USCV) estimated that the likelihood of the discrepancy between the combined state exit polls and the official state results occurring by chance was about one in ten million.

A proper combined likelihood of such a large discrepancy in both the national and state polls would multiply those two numbers, to give a result of one in ten trillion.

Although the exact number can be and has been computed in different ways by different investigators, nobody, including Warren Mitofsky, disputed the fact that the likelihood of this discrepancy occurring by chance is so small that it should not even be considered.

In 2008 the exit poll discrepancy was considerably smaller than in 2004, but it was still well outside the margin of error. I won’t calculate an exact number, since we don’t have all the data yet. But it’s safe to say that the difference is very unlikely to be explained by chance alone.

The fact that pre-election polls [1] provided an estimate very similar to the exit polls in 2008 (The Obama lead was a little bit less in the pre-election polls, but it was surging upwards in the last couple of days, so probably the two were about equivalent) makes it even more likely that they were both accurate.

So that leaves two possibilities: Exit poll bias (and pre-election poll bias as well) or impaired election integrity – that is, election fraud.

Then why did Obama win if election fraud was committed?

First of all, let me say that I don’t know for a fact that election fraud is the primary explanation for the exit poll discrepancies, either this year or in 2004. The issue was extensively investigated in 2004, and the results were not fully conclusive either way.

Two possible reasons why the results were not fully conclusive were that:

1) Independent voter activist organizations were not provided access to all of the raw data, and
2) Nobody was provided access to the “proprietary” voting machines.

But let’s assume for a minute that election fraud in general, and programming of electronic voting machines to switch votes to the Republican candidate was the major reason for the exit poll discrepancies in both elections. Why then did Obama win, if the Republicans had the capability of committing that kind of election manipulation?

The answer to that is that their election fraud capabilities are not infinite. Both pre-election polls and exit polls showed Obama winning in 2008 by a much larger margin, especially in critical (and formerly red) swing states, than John Kerry in 2004. Kerry ended up with only one state (Ohio) that was very close and would have given him an electoral victory.

Obama, on the other hand, would have won with any ONE of a number of formerly red states, all which showed him with both exit poll leads and pre-election poll leads or virtual ties (including Virginia, Ohio, North Carolina, Indiana, Missouri, Colorado or Nevada.

So when I saw Obama with huge exit poll leads in VA and OH (of 9 and 8 points respectively) and good leads in NC and IN as well, it seemed highly unlikely to me that such large leads could be overcome with election fraud.

Part of the reason for that belief was the Democratic performance in Congressional races in 2006. There were indeed substantial exit poll discrepancies in those elections – similar to what was seen in the 2004 Presidential election [9]. But if Republicans had unlimited capability for election fraud, then why did they allow Democrats to take over Congress in 2006?

And furthermore, this issue has been given a lot of attention since 2004. Obama’s leads in both the pre-election and exit polls in critical swing states this year were so great that the theft of a third consecutive Presidential election could have been too much for American citizens to bear. Even if the exit poll discrepancy this year had been substantially greater than in 2004, Obama still would have won. The theft of this election would have been much more difficult to swallow than the theft of the 2004 election.

Why this issue is so important

Like all other activities, some sort of independent monitoring is needed to ensure that election processes are conducted fairly. To settle any of a variety of disputes in our country we have courts of law and investigators to gather evidence. Our sporting events require referees. Elections are the backbone of our democracy. Without fair elections we have no democracy.

The central process of our elections is the counting of our votes. Yet we now have electronic machines that count our votes out of view to American citizens – in other words, in secret. That is not acceptable for a democracy. As noted above, a large discrepancy between exit polls and the official vote count means either exit poll bias or election fraud. This is not a matter of making political points, or rubbing in a sound Democratic victory in the faces of our opponents.

Exit polling is considered a standard practice [10] for monitoring elections, and it is especially important when vote counts are conducted electronically, with no paper trail. There is a good reason for this: Large discrepancies between exit polls and official vote counts provide an important warning sign regarding the integrity of elections.

In the presence of large exit polls discrepancies, there is no way to know whether or not extensive fraud has been committed without an extensive investigation, including access to the voting machines. After three consecutive national elections manifesting large exit poll discrepancies well beyond the margin of error, and all in the same direction, it is way past time that we find a way as a nation to ensure that our elections are conducted fairly.

Counting our votes in secret has no place in a democracy – especially when those doing the counting are heavy contributors [11] to one of the participants. In the next few days, weeks, or months the EDA will be conducting a thorough analysis to see what we can find out about the 2008 presidential election exit poll discrepancy.

We won in 2006 and 2008 because we had large, virtually fraud-proof leads. Had those elections been close, as in the Presidential elections of 2000 and 2004, we probably would not have won. Look what happened when we lost those elections. We just can’t allow that to keep on happening.