Carter-Baker Exit Poll Recommendation Would Hide Evidence of Vote Count Errors

The National Election Data Archive disagrees with the Carter/Baker Commission exit poll recommendation and urges instead, in the strongest possible way, increasing public access to exit polling data and methodologies to better monitor the accuracy of our elections. In the context of the questionable practice of altering exit poll results to match election results, this Carter-Baker recommendation would deprive the public of knowing the true unaltered exit poll results and eliminate a measuring stick against which to gauge the validity of vote counts.

(PRWEB) September 22, 2005 -- The National Election Data Archive (NEDA) is a nonprofit organization of statisticians and mathematicians devoted to the accuracy of U.S. vote counts.

We commend the Carter-Baker commission for recommending that Congress require a voter verifiable audit trail because routine independent random audits of the voter verifiable records would go a long way toward ensuring the accuracy of U.S. election results.

We believe, however, that the Carter-Baker commission report (http://www.american.edu/Carter-Baker/) erred in recommending that news organizations delay the release of any exit poll data until the election has been decided". This recommendation is ill-advised for the following reasons:

* In the 2004 presidential race, the election pollsters, Edison/Mitofsky (E/M), altered their data to match the reported vote tallies after the official 2004 results were in. (See E/M report http://www.exit-poll.net/election-night/EvaluationJan192005.pdf pgs 5 and 20.) Viewed in the context of this questionable practice of altering exit poll results to match election results, this Carter-Baker recommendation would deprive the public of knowing the true unaltered exit poll results and eliminate a measuring stick against which to gauge the validity of vote counts that could possibly lead candidates to request recounts.

* Exit polls have historically been accurate, in both the USA and abroad. Wherever paper ballots are hand counted, they are typically within 1% of the exit polls, and this was the case in the USA in 2004. When monitoring foreign elections, American officials rely on exit polls to determine whether the elections were fair or not.

* In the USA, with the increased use of electronic vote counting, driven by software which is confidential and known only to a handful of individuals who are primarily of one party, and the lack of independent audits to check accuracy, U.S. presidential vote counts have ceased showing strong correlation to exit polls in recent years.

* Virtually all of the thousands of glitches" reported by voters using electronic voting machines favored the same partys candidate. This is statistically improbable, has nothing to do with exit polls, and suggests that the counting software may have been inaccurate.

* Scientific analysis by the National Election Data Archive of the exit poll data suggests high probability that there were errors in vote counts in the 2004 presidential election. (See http://electionarchive.org/ucvAnalysis/US/Exit_Polls_2004_Edison-Mitofsky.pdf and http://electionarchive.org/ucvAnalysis/US/exit-polls/USCV_exit_poll_analysis.pdf)

If we told our votes to a man behind an opaque curtain who recorded our votes and cast our ballots for us, and then put two individuals with party ties in a room to count our votes and tell the rest of us who won, few among us would trust the results. Yet this, in essence, is what we have done with our electoral process when we cast votes electronically and allow a handful of programmers to count them.

For these reasons, The National Election Data Archive disagrees with the Carter-Baker Commission exit poll recommendation and urges instead, in the strongest possible way, increasing public access to exit polling data and methodology to better monitor the accuracy of our elections. Further, NEDA recommends public release of all detailed election data at the precinct level broken out by vote type so that unusual patterns that would be produced by vote count errors may be detected and American democracy protected.

Please see http://electionarchive.org/ucvAnalysis/US/election_officials/Audits_Monitoring.pdf for common sense recommendations for increasing the accuracy of U.S. vote counts.

###