CT: How Many Errors Can You Find In This Story?

Update: Cross posted at MyLeftNutmeg. See some of the comments there, especially Tessa’s describing obtaining election results on election night in Milford. <read>

ConnPost has an article on errors in Shelton on election night. But we find other possible inaccuracies in the story as reported: State: Shelton vote snafu ‘human error”, <read>

Shelton’s arithmetically challenged voting officials snatched away a local victory from Democratic congressional challenger Jim Himes a week after initial results indicated that he won the city, state officials have determined…

It didn’t get straightened out until Nov. 13, nine days after veteran U.S. Rep. Christopher Shays conceded that Himes had won the Fourth Congressional District race representing 19 southwestern Connecticut communities.

Bysiewicz said there was no political malice involved, a fact that the feuding local voter registrars — Democrat John “Jack” Finn and Republican Peter R. Pavone — agree upon. With a lingering controversy over an incorrect result on a local ballot question sharply dividing the two registrars, both Finn and Pavone say it was strictly erroneous tabulation that initially had Himes winning in the part of the city that’s in the Fourth District.

While the initial results had Shays with 7,114 votes and Himes with 7,632, after Pavone and Finn performed the recount, Shays had 7,668 and Himes had 6,744…

Finn said, “A mistake could have come from a person reading the number to the person on the computer. It had to be an error putting numbers into the computer.” He noted that Shays’ absentee ballots were also initially omitted.

We are pleased that the error has been corrected and that it did not change the results of the race. It would be even better if the registrars could manage to get along.

This was not the only error found in reporting results in Shelton:

“The discrepancies seem to be in the congressional races where there were cross endorsements,” Bysiewicz said. “Shelton’s one of those weird, split towns where there are two districts.”

“I definitely think it was human error, a transcription problem,” recalled state Sen. Dan Debicella, R-Shelton, who won re-election that night, but whose numbers also changed over the week and a half it took to agree on a final total…

There were also transcription errors when election officials dictated results that were typed incorrectly onto city spread sheets.

Bysiewicz believes that initial miscounts on absentee ballots was another problem…

“Arithmetic mistakes are not unusual,” Bysiewicz said, noting that her staff even found a mistake in the turnout percentage of the finalists for her “Democracy Cup” award that goes to the towns and cities with the highest Election Day turnout.

“Avon said they had 96 percent, but when we when crunched the numbers ourselves they were wrong and New Hartford ended up being the winner,” she said.

There is also plenty of confusion about the dual endorsements:

Part of the problems, Bysiewicz said last week, was that Himes was cross endorsed by the Working Families Party, so he appeared on the ballot in two places.

Some voters filled in their ballots in both spots and in those cases, if the tabulation machines did not reject the ballots, the votes were given to Himes on the Working Families ballot line.

“It’s not in [state] law, but it’s our advice to count double votes for Working Families, or whatever the cross endorsement is, because it’s up to us to help the minor parties,” Bysiewicz said.

We question the statement that “the tabulation machines did not reject the ballots, the votes were given to Himes on the Working Families ballot line”. Our understanding is that when a voter voted for the same candidate, they were counted once but then listed as UNK (unknown party) on the tabulator tape. The hand counted ballots should also have been counted that same way as UNK. We presume the Post misunderstood the Secretary.

It is also our understanding that the Secretary of the State did make a decision to not only count the UNK votes for the Working Families Party, but also to total them in the same bucket when reporting results on her website.

We are not sure if the following statement is correct:

Secretary of the State Susan Bysiewicz said a review last week of the city’s mistakes found that Shelton had the biggest Election Day tabulation breakdown among the state’s 169 towns and cities.

Our understanding is that there are still unresolved errors in Stamford, from a comment on a post on MyLeftNutmeg on the Shelton story <ref>:

Weird numbers in Stamford too… on the Constitutional Convention. The original number they sent to the SOTS office was 41,775 “no” votes. That was amended to something like 23,000 “no” votes due to an “Excel error,” I’m told. That’s a pretty big error. The original number is still posted on the SOTS web site.

If true, then perhaps it was not a “tabulation breakdown among the state’s 169 towns and cities” but made elsewhere. Looking at the Stamford numbers at the Secretary of the State’s web this morning, we see:

President Total Votes 49543 McCain 17510 Obama 31733 Nadar 289 Others 11

Congress Total Votes 47327 Shays 19735 Himes-D 26039 Duffee 213 Himes-WF 1035 Carrano 305

Question 1 Total Votes 58024 Yes 16249 No 41775

Question 2 Total Votes 49087 Yes 25679 No 23408

Looks like there is, at minimum, an anamoly of about 9,000 votes.

A big part of a solution would be for all polling place moderators to be required to fax their moderators’ returns, checklist reports, and tabulator tapes to the Secretary of the State’s Office on election night. The Secretary’s Office post the faxed images on the web, along with much more detailed keyed in results, in downloadable format. Then interested parties would be able to check the data. Of course, the remaining potential transcription error gap is hand counted ballots and hand transcribed numbers to the moderators’ reports not on the tabulator tapes.

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

One response to “CT: How Many Errors Can You Find In This Story?”

  1. The BRAD BLOG : 'Daily Voting News' For December 06 and 07, 2008

    […] CT: How Many Errors Can You Find In This Story? http://www.ctvoterscount.org/?p=1089 […]

Leave a Reply

You must be logged in to post a comment.