Friday, November 10, 2006

Why we need more exit polls

BradBlog has published what should be a blockbuster story about the latest congressional race in CA-50, the Francine Busby (Democratic challenger) vs. Brian Bilbray (Republican incumbent) match.

As you will recall, this is Duke Cunningham's old seat. The first election was highly controversial, since the voting machines were allowed "sleepovers" in private homes. The same controversy affected the Nov. 7 vote.

Officially, Bilbray won, 53% to 43%. However, Zogby conducted an exit poll in this district, and discovered a six percent discrepancy between the "actuals" and the exits.

There's more. Much of what you will read after the jump applies to our earlier essay on the Virginia and Florida-13 races.

(To read the rest, click "Permalink" below)


Here's the full scoop, from attorney and election-watcher Paul Lehto (the guy who wrote that extremely important analysis of the Snohomish County 2004 vote). Pay special attention to the startling discrepancy between the paper vote and the compu-vote:
CA 50 EXIT POLL ESTIMATES 6-POINT DISTORTION ENABLED BY REGISTRAR’S ABUSE OF DISCRETION

Haas’ Decision To Delay Paper Ballot Count
Artificially Inflates Votes For Republican

November 10, 2006 San Diego, CA – Attorney Paul Lehto, a national expert on election integrity, today released the results of an independent Zogby exit poll of voters in the November 7 election in California’s 50th District. The exit poll shows the race much tighter than the results published by the San Diego Registrar of Voters as of the day after the election, six percentage points closer. Moreover, after adjustment for an oversampling of Bilbray voters, the adjusted poll shows Democrat Francine Busby well within the margin of error, making the race uncertain in actual outcome. Paper based votes yet to be counted favor Francine Busby, such as paper ballots obtained at the polls, which constitute 6.4% of all voters and favor Busby 58% to 39%.

The poll also contains important findings about the relationship between different forms of voting and voter demographics, revealing a tendency for Democrats to use paper ballots versus other forms of voting such as electronic machines. In violation of state and federal law, and despite lawsuits filed to stop the process, San Diego Registrar of Voters Mikel Haas delayed the paper ballot count for paper ballots obtained at the polls by fully qualified voters, preferring to have other forms of ballots have been counted preferentially first, attempting to justify this second class service to paper-based voters by stating “this is an electronic election” when in fact the public is legally allowed a choice of either technology.

In a midterm election marked by concerns about the security and integrity of electronic voting, the delay essentially eliminates paper ballots from the early results typically publicized on election night, and thus eliminates Democratic votes from the consideration of what the election means for candidates, for the public, and for the perceived competitiveness of districts for several news cycles if not permanently, and therefore leads to distorted news headlines and interpretations. This can only be avoided if the media realizes that in situations where choices of voting technology are available, it is becoming increasingly common for voters of different parties to prefer different technologies.

According to Lehto, “Mr. Haas’ abuse of discretion stems from the current nationwide bias by election officials toward electronic voting which, in San Diego’s case, skewed election day results toward the Republican candidate. The resulting bias in media coverage caused irreparable harm to the other candidates. This preference flies in the face of serving actual voters, which the Zogby poll establishes that over 80% of them prefer a voting system based on the ability of the public to witness and observe vote counting (like paper balloting) and not one based on invisible electrons and trade secrecy (touch screen voting). Delaying the count for one class of voters that’s more Democratic but not for another class that’s more Republican also creates 'second class ballots' and 'second class voters' who don’t get considered by the media and politicians on election night and the day after, which is outrageous and unconstitutional. It is very clear in the law that each voter is entitled to 'equal dignity' under the Constitution.”

Given the deep mistrust of electronic voting by people across the country, the Registrar’s office had ample reason to anticipate that many San Diego voters would reject electronic voting. In fact, the exit poll shows that 12.7% either had a problem in voting or know someone who did, a very high percentage for what should be a simple process that’s been practiced for centuries. In fact, media coverage before the election cited Haas’ admission that an insufficient number of absentee ballots were printed. Poll worker training was reportedly focused on electronic voting, and Haas imposed an arbitrary limit on the number of paper ballots available, despite widespread reports of problems due to problems with electronic voting machines. In other jurisdictions, such as Maryland, problems with Diebold voting systems created chaos during the September 12 primary and led the governor to push to scrap electronic voting.

In a related development, a petition which names Haas as respondent, alleges violations of due process and equal rights in voting in a case called Peterson v. Haas. (Ed Note: Details on that complaint here and here.) By limiting the number of paper ballots available and delaying the count of paper ballots, Haas violated requirements issued by Secretary of State Bruce McPherson. McPherson had issued a letter on October 3 to all county registrars which provides that all county officials shall have an adequate supply of paper ballots, available at the voting locations for use in the event of a loss of the ability to use electronic equipment, or if a voter chooses not to vote on electronic equipment. Haas failed to do so even after clear evidence of failure presented itself, preferring instead to “transfer” votes from photocopies to official ballots after being voted by voters, a process voters certainly never agreed to, and which violates laws preventing third parties from assisting voters with voting unless necessary. It is unknown how many mistakes or intentional errors were made in transferring the voters’ votes onto new ballots.
A six percent error is odd in a predictive (that is, pre-election) poll. In an exit poll, that sort of "glitch" is downright suspicious.

1 comment:

Anonymous said...

It's important to note that the gap could narrow as there are still uncounted votes in large numbers, even though the REgistrar claims 100% of precincts reporting....

Nevertheless, here in San Diego County, in the 50th Congressional District Busby/Bilbray race,
based on the Zogby exit poll we had performed, one of the following two very bad things
MUST be true, logically:

(1) EITHER there is a very large discrepancy in the Zogby
exit poll vs posted results so far, which means a suspicious election
or a defective poll, and 95% of Zogby's polls are within 1% of
actual result, OR ELSE

(2) in light of the 180,000 ballots still to be
counted in San Diego County, an unknown percentage of which are from
the 50th, the race will tighten up considerably over results reported
so far, consistent with Zogby exit poll data suggesting that Democrats
disproportionately chose paper ballots and Republicans
disproportionately chose electronic ballots.(the main point of my press release) This is the "SECOND CLASS BALLOTING SYSTEM"
where paper balloters are dealt second class service that doesn't get counted or considered by media headines in the days following the election.