Sunday, October 26, 2008

Is the difference between exit polls and results due to faulty exit polls or Election fraud? Probably a bit of each

The electoral ‘irregularities’ uncovered by Palast and Harris might explain some of the difference between the 2004 exit polls – in which Kerry was leading – and the final result – a victory for Bush. Professor Steven F. Freeman of Pennsylvania University examined US Presidential election exit polls carried out by Edison Media Research and Mitofsky Polling on behalf of a media pool including the Associated Press. These show exit poll discrepancies in every key swing state in the US Presidential election, all favouring Bush, including 4.5% in Florida, 6.7% in Ohio and 9.5% in New Hampshire. (1). Democratic party Congressman John Conyers of the House Judiciary Committee requested that Mitofsky release the raw exit poll data on which Freeman’s study was based. Mitofsky refused to do so on grounds of commercial confidentiality (2,3)

President Bush and some other senior republicans immediately called for an inquiry into the discrepancy between the exit polls and the results in the 2004 elections in Ukraine. (Exit polls that differ greatly from the final results are usually seen evidence of election rigging in other countries). The Chairman of the Republican National Committee simultaneously asked the media not to report exit polls in US elections in future, with the discrepancy in the US explained as due to the exit polls, getting it wrong (4). Now that’s some impressive double-think. The Soviet Union’s rulers would have been envious of getting people to buy that one.

The American exit polls were modified after they were taken. The ‘explanation’ that it was ‘standard practice’ to ‘weight’ the polls by the final results is more an admission that the exit polls were massaged after they were taken. Even its originator – Democratic pollster Mark Blumenthal- later largely agreed with Freeman that this was not good practice (5).

The Bush administration also condemned the Ukrainian elections on the grounds that international observers had criticised them. The International Herald Tribune meanwhile reported that:

‘international monitors at a polling station in southern Florida said Tuesday that voting procedures being used in the extremely close contest fell short in many ways of the best global practices.

The observers said they had less access to polls than in Kazakhstan, that the electronic voting had fewer fail-safes than in Venezuela, that the ballots were not so simple as in the Republic of Georgia and that no other country had such a complex national election system…..

…. Variations in local election law not only make it difficult for election monitors to generalize on a national basis, but also prohibit the observers from entering polling stations at all in some states and counties. Such laws mean that no election observers from the organization are in Ohio, a swing state fraught with battles over voter intimidation and other polling issues.’(6)

Yet strangely the courts found that it was legal for Republican and Democratic party representatives to be present in the polling stations to challenge voters on whether they had the right to vote there (7).

Republican congressman Jeff Miller of Florida told the international observers they should “get on the next plane out of the United States.” (8). So international observers’ reports are evidence of vote rigging in Ukraine, but not in the US.

Some experts on polling and statistical analysis have pointed out that there could be other explanations for the exit polls being wrong and that exit polls aren’t always accurate.

They’re right – there could be other explanations and I’ve previously been wrong in following Palast and Robert F Kennedy Jr. in talking as though exit polls are always right
and as if differences between them and final results were cast iron evidence of fraud.

One convincing alternative explanation, cited by Mark Blumenthal, is that many Republicans believe the media and polling organizations to have a ‘liberal bias’ and so fewer Republican than Democrat voters might have been willing to answer pollsters’ questions on who they voted for (9). Also older voters may be more likely to vote Republican and less likely to answer pollsters' questions. These are both convincing theories with some evidence to support them and might partially explain the exit poll discrepancy, but they don’t disprove Palasts’ or Harris’ claims either.

The fact that Harris and Palast can provide solid evidence of real voters being illegally removed from electoral registers in large numbers and that they and Verified Voting also provide evidence of several other methods of election manipulation having been used lends weight to their claims about the reasons for the difference between exit polls and final results.

Some of the supposed proof that the 2004 election results weren’t manipulated isn’t proof at all either. For instance Blumenthal quotes the Democratic National Committee’s 2005 report on the 2004 elections, which said:

‘Strong similarities at the precinct level between the vote for Kerry (instead of Bush) in 2004 and the vote for the Democratic candidate for governor in 2002 (Hagan) present strong evidence against the claim that widespread fraud systematically misallocated votes from Kerry to Bush. In most counties we also observe the pattern we expect in the relationship between Kerry's support and other precinct-level factors: Kerry's support across precincts increases with the support for the Democratic candidate for Senator in 2004 (Fingerhut), decreases with the support for Issue 1 and increases with the proportion African American.’ (10)

This is a strange claim unless the DNC were assuming that Palast was claiming it was only the 2004 election and only Presidential votes in it that were being tampered with. Of course he wasn’t. He provided evidence that all elections were being tampered with, so similarities between the counted votes for Democrats running for different offices or in different years in the same districts don’t disprove his claims at all. Attempting to rig only Presidential votes would be practically impossible except where electronic voting machines were used.

The DNC are also assuming all the allegations of electoral manipulation involved people taking Kerry votes and giving them to Bush. That may well have happened with electronic voting machines, as Bev Harris and John Hopkins University found it would have been entirely possible with the systems used in 2004 – but much electoral manipulation involved other methods – such as simply not counting provisional ballots (since most were cast by people from a mostly Democratic voting demographic who had been removed from the electoral register by Republicans or firms working for them) or providing less and older voting machines in mostly Democratic districts in order to ensure far more Democratic than Republican votes would end up discarded as ‘spoilt ballots’.

De Ja Vu 2008?

Much of the media are now cautioning that polls may be misleading and even that the exit polls may well be wrong due to the 1980s ‘Bradley effect’ (more white voters saying they’ll vote for a black candidate in polls than actually do). In fact the Bradley effect has not existed for any black candidate from 1996 on (11).

So if you see the exit polls being very different from the polls come November 4th and the results being very different from the exit polls – and all in McCain’s favour – there’s a reasonable chance that it might not just be faulty polls and faulty exit polls. Something might have also gone badly wrong with the electoral registers, the lack of a standard, reliable voting system in every district - and the counting of the votes.

(1) = Professor Steven F. Freeman – CV - -

(2) = Professor Steven F. Freeman 10 Nov 2004 , ‘The Unexplained Exit Poll Discrepancy’, and

(3) = House Committee on the Judiciary – Democratic Members – Correspondence - Text of Letter from Rep. Conyers to Warren Mitofsky of Mitofsky International Requesting the Release of Exit Poll "raw data" (12/3/04) ,

(4) = House Committee on the Judiciary – Democratic Members – Correspondence - Mr. Mitofsky Response (12/7/04) ,

(5) = Television Week 5 November 2004 , ‘GOP Wants News Organizations to Abandon Exit Polls’ ,

(6) = Mark Blumenthal – Mystery Pollster Blog – Exit Polls -

(7) = International Herald Tribune 03 Nov 2004 ‘Not a simple election, global vote monitors say’,

(8) = USA Today 02 Nov 2004 ‘Court clears way GOP reps to challenge voters' eligibility in Ohio’,

(9) = International Herald Tribune 03 Nov 2004 ‘Not a simple election, global vote monitors say’,

(10) = Mystery Pollster
Demystifying the Science and Art of Political Polling - By Mark Blumenthal
06 July 2006 ‘Is RFK, Jr. Right About Exit Polls? - Part IV,

(11) = San Francisco Chronicles 21 Oct 2008 ‘Many think 'Bradley effect' won't hurt Obama’,

No comments: