Saturday’s NY Times printed a story about the published results of the New York Democratic Presidential Primary having numerous errors.

While the story made it clear to anyone who read it closely that the errors would actually have no bearing on the allocation of delegates, the inevitable has happened. Numerous conspiracy theorists got to work immediately and claimed that this was another case of an election being stolen.

I was going to post something about this but late on Saturday, I received an e-mail from Doug Kellner, the Democratic Co-Chair of the New York State Board of Elections that explained how the votes are counted in New York much better than I could. Here is Kellner’s report –

Sam Roberts' report in this morning's New York Times is really old news to any experienced New York City politico. At the close of the polls, the election inspectors manually copy the numbers from the voting machines onto the "canvass report," which is prepared in triplicate. One copy goes to the borough office of the Board of Elections together with all provisional and emergency ballots. A second copy is maintained by the local police precinct. A third copy is used by a clerk at the local police precinct who types the numbers into the police department main frame computer, which is immediately available to the Election News Service, a local media consortium, who eventually receive the third hard copy of the official canvass report.

Everyone in New York knows that the election night numbers typed by the police clerks are very unreliable and filled with typographical errors. Most of the local Democratic clubs have watchers present at the close of the polls who record the numbers independently. In this particular race, most clubs had both Obama and Clinton supporters who freely shared their numbers. It has been my experience that the numbers collected by the local political organizations are much more accurate than the numbers distributed to the media through their election night consortium.

The NY Times did accurately note that in New York Clinton had the first column and Obama had the fifth column, with candidates who had dropped out in the second, third, fourth and sixth columns, who often did receive zeroes on the voting machines. Therefore, it was a relatively easy typographical error for the clerk at the Police Department to put a zero in the Obama column, which occurred in a couple of precincts. It has been my experience that almost all of the errors are police clerk typos, rather than transcription errors by the election inspectors who prepare the canvass reports.

The silver lining is that because everyone knows that these election night numbers are filled with typographical errors, no one I know (except the media!) ever relies of these numbers. New York has a mandatory, 100% recanvass, which both election officials and the political candidates generally take very seriously, even when the races are not close.

Although I don't have the exact percentages for this presidential primary, absentee ballots are typically 3.5% of the total and, in New York City, valid provisional ballots are about 2% of the total. It is extraordinarily rare to have any argument over the numbers ultimately certified because the candidates’ watchers and the board officials generally exchange their numbers a make an effort to reconcile them. The candidates’ watchers have access to the lever voting machines, which are removed to the borough voting machine facilities the day after the election, and may look at the paper ballots, during both the initial count after the election and the recount that takes place eight days after the election. Absentee ballots postmarked before election day are valid if received within seven days after the election.