Lynn Landes  Lynn Investigates ...

Go back to Voting Machine Webpage

Return to Parallel Elections webpage

First “Grass Roots” Parallel Election Project

Coconut Creek, FL, March 8, 2005.


By Ellen H. Brodsky


Bill Moyers, famous journalist, researcher and problem solver has said “Most issues of importance have been solved by local people, the “Grass Roots” of the nation.


Issue: Are the voting machines accurately recording the votes according to the voter’s intent?

We are only told whether an election ran smoothly or not according to election officials which is mirrored in the media coverage.

 Definition:  The election is over. No public scrutiny and all the machines are safely back in the voting machine warehouse.


When our Supervisor of Elections is asked to explain why the voters had such difficulty voting for Kerry and the screen kept switching to Bush and other machine or ballot problems, we are told that it was the voter that should have reviewed the summary screen, or have been more careful.  But what about ES&S  programming the machine so that this does not occur?  We are told the machines have been thoroughly tested for accuracy under stringent FL Law requirements. When asked has the Supervisor or her staff examined the election software to make sure it is accurate, we are told to speak with their lawyer. In all seriousness, I even found one voter who tried to press Bush and got Kerry!


 The people of Coconut Creek are not stupid.  Corporately owned machines with secret software and no voter-verified paper record are doing all the voting, and there has never been any Statistical Validation to compare the voter intent with the machine results. The voters of Precinct 12F in Coconut Creek were tired of being blamed for machine malfunctions and forced to vote involuntarily into a black box. Over 67% of the voters in our study had absolutely no confidence or evidence that their votes were being accurately recorded according to their intent and welcomed the idea of doing their own testing. They drew the line in the sand and said this is the people’s election, from now on no more “Faith Based Voting” and took it upon themselves to exercise their patriotic duty by checking the accuracy of their voting machines. So began the birth of the first ever “Grass Roots” Parallel Election Project!


This project was so simple yet proved more accurate and more powerful than any exit poll and provided hard evidence as to machine programming problems and clarified voting irregularities.  Since the 2000 election there has been little faith in our democratic elections.  Touch screen machines with no voter verified paper ballot have made reconciliation and recounts virtually impossible as there is no way to independently count the voter intent. The software is so secret that even the Supervisor of Elections has no access. The audit log will only tell you that a ballot was cast but not what choices the voter made. Unscrupulous elected officials and Corporate Vendors have used the requirement of secret elections to make the whole election process secret to the voters and yes even our elected officials who are supposed to administer our elections.  The majority of participants said “Wow, what a great idea, why hasn’t this been done a long time ago?” Some of the candidates said this should be done in every precinct in Coconut Creek. One of the Commissioners asked me why wasn’t this done in precinct 11F? Other cities have expressed an interest in doing this in the 2006 Governor’s race. I believe this is the best method of checking the accuracy of the voting machines during an election by empowering the citizens themselves. It restores faith in the election process by the voters directly marking, and yes, casting their ballot the old fashioned way in a ballot box and counting their own votes. It finally gives them evidence if the machines are accurately recording their votes. Voting is almost a lost art, as most Americans have forgotten what voting is all about. There is no better way for a Supervisor of Elections to restore confidence that the votes are being counted accurately than to incorporate the Parallel Election Project as an integral part of all elections. It insures that the correct candidates or issues are truthfully being elected. If there are major variances in the Parallel Election and the machine totals there is hard evidence to challenge the election results. If there is evidence of bad programming an independent audit should immediately be ordered.


One question asked is it legal?  We had no access to the software, and it was a grass roots test comparing the results. It was a test only. It has never been done before so there is no precedent! The voter officially only voted once. It may or may not have been if I had used the official ballot or tried to copy the official ballot. But I rewrote and designed the ballot to include the candidates and issues mostly in my own words. Of course some words could not be changed like the names of the candidates, or yes and no. I made a point of calling it the Parallel Election Ballot and again on the bottom told voters that this was not the official ballot.

Is it illegal for the citizens to do their own independent testing of the voting machines they are using?  Is it illegal for the voters to have hard evidence to use in the case of an election challenge? Is it illegal for a citizen to pledge that they voted a certain way on a test ballot by signing their name because they want to be contacted in the event of a challenge? Besides it is not mandatory to sign the Parallel Election Ballot.

In the District 91 Race back in January, 2004 there were 134 under votes and 12 votes separating the top two candidates, what other legal way could the winner have been determined? And what about allegations of vote rigging and machine tampering in Ohio and Florida in November 2004?  If our Democracy is to survive, the citizens must verify their own elections!  Wouldn’t it be great if Supervisors of Elections recognized the Parallel Election Project as standard practice in all elections in the future?


Preparation: Balloons, Ballots and More…


I designed and printed 10,000 ballots and 10,000 handouts. I scouted each precinct location to find the best location to setup. I had to arrange enough chairs, tables and supplies for each precinct. I had to be prepared for the weather since we had to be outside the polling site which was predicted to be rainy, cold and very windy. I purchased a tent and tarp just in case.  I went to a Party Store and rented a helium pump to inflate loads of colorful balloons, sorry no coffee and donuts this time.  I was careful to indicate that this was the Parallel Election Ballot and not the official election ballot. My house became full of potential cardboard ballot boxes gathered from local stores which could have served well in a VotersUnite ‘s “Blackboxvoting Tea Party.” My initial goal was to have PE Poll workers stationed at 7 Coconut Creek Precincts. I needed to be prepared to have enough ballots and handouts for every registered voter in these precincts.  I made special arrangements with the Coconut Creek Government Center, Precinct 12F to let them know of what we were planning to do on Election Day. We had to setup 50 ft. from the polling site as designated by the Poll Deputy. (See Handout and Ballot attached.)

I distributed flyers to many communities in Coconut Creek describing what a Parallel Election was and asked them to participate on Election Day. I contacted the Sun-sentinel and Miami Herald and received a call from one who interviewed me about the project.

The most important element for success is having strong reliable manpower and planning for all eventualities. I decided my role would be Chief Elections Clerk and Roving Supervisor to augment supplies and fill in where needed.  With large precincts of 1,000 to over 3,000 registered voters such as in Broward County you must have at least 2 people at all times stationed at each precinct who are physically agile. Your location must be such that all voters entering and leaving the precinct must pass by your station considering the 50 ft. rule. Each location must have leaders capable of taking command should one Poll worker fall. For example, one location in Wynmoor, a senior community, that had 4 precincts would have been ideal except paid supporters of the Gambling referendum were intimidating and fighting with my PE Poll workers mostly elderly, for the prime location outside the polling site, making it impossible to do their job. They claimed our PE Poll workers were interfering with the election. (Remember the recounts in Miami in 2000?) A strong PE leader and enough able manpower could have remedied the situation, won the claim, relocated or carried on.  I decided as a result, to consolidate my Parallel Election effort to one precinct. I had all the volunteers assigned in shifts manning the polls with me at Precinct 12F. By consolidating the Parallel Election effort to one precinct, I was guaranteed reliable manning of the polling site from 7 am when the polls opened to 7 pm when the polls closed to accomplish a controlled real-time Parallel Election. It was also helpful that one of the PE Poll workers was a respected ex-Commissioner and that I had secured permission from City Hall. Also, some of the Commissioners had told me they wanted this done in every precinct in Coconut Creek although they were personally unable to help in this election.


“In large precincts such as in Broward County you must have at least 2 people at all times stationed at each precinct. Your location must be such that all voters entering and leaving the precinct must pass by your station. By consolidating the Parallel Election effort to one precinct, I was guaranteed reliable manning of the polling site from 7 am when the polls opened to 7 pm when the polls closed to accomplish a controlled real-time Parallel Election. “


What happened?


Having been an Election Clerk since the use of Touch Screen Voting, I conducted the election the same way it is done in the Precinct. Before the first voter voted we opened the Ballot Box to demonstrate that it was empty and no ballots had been cast. The machine equivalent would be the printing of the zero tapes showing no public count.  I announced to my PE Poll Workers and the public at 7 am that the Polls are now open.

We kept a sign-in list like the Precinct Registers to compare numbers of ballots cast with voters who participated. We gave everyone who passed our controlled area on the way to the Precinct a handout describing the Parallel Election Project. We asked them to vote in the official election and told them that after they voted and exited the precinct we would give them a Parallel Election Ballot, which we did. We asked them to vote exactly the same way they voted on the voting machines not how they would have liked to have voted. We let the voters know we were checking the accuracy of the voting machines. We got all manner of feedback as to their voting experience. The most striking problems were the absence of the Commissioner Race on many ballots, and many of those that had voted the race said the type was very small in comparison to the gambling referendum. I instructed the voters who were missing the Commissioner Race to write this on their PE Ballot, which many did. I designed the ballot so that the voter could sign the bottom portion which we told them could be used as direct evidence if there was a need for an election challenge. I also offered them the option of not signing if that made them more comfortable about participating. After all, the aim was getting as many people possible to participate and to test if the voting machines were accurately recording their votes.


186 voters passed our table by the main entrance. 125 voted the Parallel Election Ballot, 61 did not want to participate. We opened the Parallel Election Polling site at exactly 7 am the same as the Official Polling Site. We gave voters a handout describing the Parallel experiment and asked people to vote in the official election and then vote in the Parallel Election exactly the same way they voted inside. We had the voters sign a register to indicate that they participated. There was a section on the bottom of the ballot for signing in the event of an election challenge; this area was sealed with a glue stick by the voter and poll worker if the voter needed help. This made the voters more comfortable who were concerned about the secrecy aspect, however, most were not concerned with this issue who participated and welcomed any follow-up and even wanted to be contacted if there was an election challenge. The voter did not have to sign the ballot to participate. Everyone who participated did so because they thought it was a great idea and understood it was an important experiment. They were excited to finally be able to check the accuracy of the voting machines. 67.2% an overwhelming majority of those that passed our setup in order to vote enthusiastically participated.  Most people were really excited to participate and felt this type of testing by the people should have been a long time ago.


The only problems I had during the day at 12F was the possibility of being closed down if I crossed the 50 ft. line by the Poll Deputy. A Supervisor from the Dept. of Elections was monitoring us all day from the Parking Lot. Finally he came and asked me for a sample ballot and left without a problem. At one point we had to clarify with the Poll Deputy that we were telling people to vote in the official election first and then vote in the Parallel Election.   That was a close call!


The marking and casting of the voter’s ballots continued joyously throughout the day. One by one we got feedback as to problems the voters encountered in ballot design and casting of the ballots. At 7 PM I announced to my PE Poll workers and the universe that the Polls were now closed.


Counting the Ballots.


We counted the ballots and the Sign-in sheet record and found that we had 125 ballots cast, 125 signatures and 61 non-participants who passed our controlled study. The majority of the 61 that did not participate said the machines were accurate, or computers are 100% accurate or to a lesser degree they didn’t want to sign anything. A few passed by as though they were in another dimension without acknowledging we were there, making a beeline to the parking lot.


I went to the Precinct and watched them count their signatures and print the poll tapes and wrote down all the data from the totals tapes to reconcile with our test. There were 270 people who voted the official election ballot. 84 of these voters did not pass by the main entrance to vote. The weather was bad and these people may have worked in the Government Center. It would have been great to have all these voters participate; however, it was not necessary to get an accurate sampling as you can see below.


Exhausted,  I packed everything into my car and we went to a local diner to have dinner and count the ballots on a tally sheet. There I was met PEPs  Joyce, Roz, Harold and two new PE Poll workers, Erika and her 7 year old daughter Emily.  Counting of the votes was the most joyful part of the whole election. Joyce separated the ballots into 2 piles one for each candidate and did the reconciling of the counts, Erika held up the ballots one by one for all to see and put the markings in the tally book as sweet Emily delightedly called out the name of each candidate for all of us to witness. For each candidate we drew a vertical line for 4 votes and the fifth vote was a diagonal line across. We entered the tally for each candidate in the tally book. We combined the pile again and repeated the process by putting the yes votes for the gambling referendum in one pile and no votes in another pile indicating the votes on a tally sheet. We repeated this process to determine which ballots were missing from the Commissioner Race. A missing (no vote, machine under vote equivalent) Commissioner Race pile and a voted Commissioner Race pile. Each counting procedure had to account for the 125 ballots cast. Our no vote study explained the machine under votes, since the voters had testified that that race was missing on their ballot as pledged and written on many ballots. It should also be noted that even though many voters complained about the small type of the Commissioner Race on the Ballot, they still were able to vote that race. We precisely isolated and analyzed the ballots containing the missing race. We found that 56.3% of the missing (under votes) for Commissioner also voted no on the Gambling Amendment, 43.7% of the missing (under votes) voted yes on the Amendment.  It is interesting to note although not a factor in our study, the Pro Gambling group had paid for the election. It is also interesting that there were 16 missing Commission Races and only 1 under vote or no vote for the Gambling Referendum and we were lucky to have caught this voter in our test. The voter testified that the Gambling referendum was not on her ballot.


Commissioner Race


Parallel Election Results.                              Precinct 12F Machine Results.


Total Public Count: 125                                    Total Public Count: 270


Total Votes Cast:     109                                   Total Votes Cast:     233


Pat Duaybes:      19 votes      17.4%                 Pat Duaybes:              43          18.5% 


Marilyn Gerber:  90 votes      82.6%                 Marilyn Gerber:        190          81.5%


No Commissioner Race on Ballot: 16   12.8%       Under votes:  37          13.7%

(Voters testified that the Commissioner

Race was missing from their Machine Ballot)





Gambling Referendum


Parallel Election Results:                              Precinct 12F Machine Results.


Total Public Count: 125                                    Total Public Count: 270


Total Votes Cast:    124                                    Total Votes Cast:    269      


Yes (For):                82      65.6%                    Yes (For):                186       68.9%


No (Against):             42       33.6%                  No (Against):            83       30.7%     


Under votes:              1           .8%                 Under votes:               1           .4%

(Voter testified Gambling Race was

Missing from Machine Ballot)



Signatures on Sign in Sheet: 125                  Signatures in Precinct Registers: 270




What we discovered was that in the Parallel Election 16 voters signed, wrote and testified that the Commissioner Race was missing on their ballot. That only the Gambling Amendment was included in their ballot. Out of 125 Ballots we had 1 voter who testified the Gambling Race was not on their ballot. Ladies and Gentlemen, thanks to the Parallel Election Project we were lucky enough to catch that voter in order to prove the reason for the only machine under vote. Less than 1% of the voters had no Gambling issue on the ballot, but between 13-14% were missing the Commissioner’s Race! And as mentioned previously the Pro-Gambling group paid for the election.


Universally without a shadow of a doubt every under vote detected in our sample proved that the issue or race was missing from the ballot. Beyond a doubt there were problems with the programming for this election in Coconut Creek Precinct 12F and Ballot Style #12.  It is clear then that the Under votes from the ES&S iVotronic Voting machines were not under votes (no votes) decided by the voters but were in fact not included on the voter’s ballot itself.  There should be an audit of the Voting Machines of Precinct 12F and all machines using Ballot Style #12, to find out what programming problems contributed to such an unacceptable error rate of missing Commissioner Races on the ballots which the machines recorded as under votes.


Incredibly, a witness to Early Voting in Coconut Creek called me before the Parallel Election who wanted to  testify that she did not have a commissioner race as well. Everyone in Coconut Creek had the same ballot style #12 whether they voted on Election Day or Early Voting.  The Parallel Election confirms that the Programming was seriously defective to have so many missing commissioner races. Certainly, if the machines had been adequately tested during the Logic and Accuracy Testing, this programming error would easily have been discovered and corrected. The 2% statewide requirement for testing of the voting machines is woefully inadequate. Every machine must be individually tested before an election for accuracy of the voter intent and accuracy of all ballot positions. It is a fact that our Supervisor of Election’s team has only ceremonial involvement with the testing and programming of our voting machines. Under the Digital Millennium Copyright Act purchasers of the voting equipment (our state and local governments) are not allowed even if they wanted to independently examine the election software and how votes are recorded.

Here the Parallel Election Project is invaluable. We the People were able to do a meaningful accurate testing of the voting machines and gather hard evidence to prove all manner of voting irregularities. We were able to detect simply and independently what our Supervisor of Elections, the State of FL and the ES&S  Team with all the supposed testing were not able to do. Voting is a three step process: Marking, Casting, and Counting of the People’s Ballots. That was what the voters of Precinct 12F did on March 8, 2005 and it was a great success!


Having represented in Broward County for awhile now, I am always searching for new ideas and initiatives from around the country in order to recommend creative options for Broward County Election Reform to our elected officials and activist community.  I have attended now a few Logic and Accuracy Testings and have been woefully disappointed by the lack of transparency of the testing process and the shockingly inadequate testing procedures.  Not only is this testing poorly or nonexistingly advertised to the public, the test itself is no more than a demo. It is clear if any testing was done it was done by ES&S Programmers in secret way before the public testing date as evidenced by the zero tapes from Early Voting being zeroed days before the Logic and Accuracy test itself and up to a week before the start of actual voting. Many of the problems that surfaced in Precinct 12F on March 8, 2005 and Nov. 2004 did not surprise me based on what I witnessed at these testings. I was thinking, there will not be another major election until Nov. 2006’s Governor’s Race. There are only 2 issues on the ballot in Coconut Creek. Let’s do something revolutionary!  I have always admired the ideas of Lynn Landes and recently stumbled across the Parallel Election idea on her website I thought it a great idea then and now even more. I said to myself, let’s give it a go!  Until legislation is passed that mandates that voter-verified ballots be counted manually at each precinct, compared against the machine results and declared the official election result, this is the most transparent and accurate method of reconciliation.  This is the most democracy enhancing, effective and downright joyous way for the citizens to mark, cast and count their own ballots while comparing the accuracy of their voting machines.  With all the lawsuits, election challenges and shenanigans by partisan election officials the Parallel Election Project provides hard evidence that could be used in a court of law. This could be the start of something big, if the citizens have the courage to take back their elections and the elected officials support them in the public interest. Everyone considered it “OUR” Project!  I want to give special thanks to all the voters that participated and the PE Poll Workers;  Harold, Charlotte, Roz, Ruth, Peggy, Emma, Bruce, Sandy, Joyce, Erika, sweet Emily and my mom who gave me some great ideas and inspired me to see this project to completion.


Ellen H. Brodsky, 954-973-2819