Problems Faced by Voters at the 2008 Presidential Primary Election

Friday, March 7, 2008
Ronald Reagan State Building Auditorium
300 South Spring Street, Los Angeles
Testimony by Judy Alter, Director, Protect California Ballots

Thank you to Debra Bowen for the new requirements to safeguard our election processes, for providing the CA Election Code requirements to count the uncounted non-partisan presidential primary votes, to the several managers in all the divisions at the LA County Elections Division who answered my questions and verified my observations, and to the volunteers from Protect California Ballots who observed many facets of the election and who contributed to this collection of reports through Work the Vote and the L.A. Election Protection Task Force.

Topics in this report:
        • voter issues,
        • poll worker issues,
        • poll monitors/observer issues,
        • chain of custody and quality control issues in Norwalk.

Source of this information: approximately 50 precincts, 65 voters and observers from all over L.A. County.

 

Voter issues:

The non-partisan “double bubble” problem was the most frequently reported issue stated by voters who did not fill in the 5 or 6 on their ballot. Many of those votes have now been counted.

Voters’ political parties not accurately listed in the roster books was the next most reported issue. Some of these voters received the wrong sample ballot. They were told to reregister.

Names were not in the roster books of long-time registered voters or not updated because of a change of address or name: another a major complaint.

Precinct change occurred frequently where voters received no notice. In a few cases the wrong address for the precinct is listed. And voters in Beverly Hills discovered that for voters who live between 200-320 on Beverly Dr., their names were not in the roster book in a precinct they should have been able to vote in.

Signs for the precincts were inadequate and voters could not find them.

Voter misunderstandings: some did not know that the American Independent party was the John Birch political party; that the Libertarian party did not mean liberty; and that NP voters could NOT vote as Republican in a primary.

Many precinct-based scanners did not work, poll workers did not know how to prepare them for use (NP ballots did not go through), poll workers did not protect the voters’ private ballot when putting the ballots into them, and voters could not see the slot in the ballot box because it was at the back of the box.

One handicapped access entrance door was closed.

Absentee ballot issues included receiving one with the wrong name on it, voters were listed incorrectly as an absentee voters, and having the AV ballot rejected, delivered by a relative, because the envelope did not have all the required information on it.

 

Poll site and poll worker issues:

Poll workers not trained to tell NP voters about the #5 or #6 bubbles. Some poll workers gave non-partisan voters democratic or AI ballots. They were not trained to explain the choices for NP voters in the primary.

Poll workers were not adequately trained to work with the precinct based scanners/readers.

Poll workers were rude to voters, and confused about voting materials and procedures such as using the provided cell phone or the adjacent precinct map. Some poll workers asked for ID from registered voters.

One inspector was not trained, one arrived at 7:00 so the poll opened at 7:50, and coordinators could not answer all the questions or handle the technical problems with the electronic machines.

Some poll sites had too few poll workers, had inadequate breaks, did not have time to update the voter list to be posted on the precinct doors, and they did not follow poll closing procedures in the recommended sequence.

Supply problems included not enough registration forms, democratic ballots and a ballot recording device; the pencil sharpener did not work and not enough ballot marking pens that worked. Replacement materials were not delivered.

Problems with the electronic scanner occurred frequently: they did not work, jammed, or the poll worker did not put in all the header cards when turning them on and so they rejected some party’s ballots.

The audio units were not used.

 

Chain of Custody issues at LA County Election Headquarters in Norwalk

The Manager of the absentee ballot voting division has not documented chain-of-custody of the absentee ballots processing (now called vote-by-mail). The ballots are first counted daily when they arrive from the post office and the final count occurs when they are scanned in, starting 5 days before the election, and tabulated on election night. Even though the ballots remain in one room until they are electronically imaged before the election, they move from area to area in a large room and many employees handle them. Batching and counting these ballots internally would reassure VBM voters that their ballots are safe and accounted for.

The large numbers of provisional ballots require processing in several sections in the elections division on several floors in Norwalk. I did not see batches of these ballots clearly marked by an id number, or date, or count in the processing sequence of these ballots. (That could be my oversight.) I did observe, however, careful quality control in the remake area with clearly written out guidelines for the temporary employees who do the remakes. And each process has quality control people double-checking each step. Thus, in spite of the high number of post election ballots received in Norwalk, each provisional ballot and every vbm ballot receives scrutiny and attention. Most of these got counted.

The election division employees double-check the signatures in the roster books and compare them to the total votes cast and counted on election night. Instead of having employees add up the total signatures on each page on an adding machine, this process is now computerized. The electronic totals of the two sources of totals are compared as quality control but no hand count is now done to check to see if the computer counts are accurate.