Venezuelan News and Analysis -                       |   printer friendly version

Venezuela: Research Team Reports 70% of the Voting Machines Were Manipulated in the 2004 Presidential Recall Referendum

By Roberto Giusti | El Universal

According to an academic team, the Yes-vote won the recall referendum by more than 5 percentage points. The interdisciplinary team did an exhaustive analysis of the data transmission system and concluded that on the 15th of August of 2004 between one and a half and two million virtual votes were added to the No-vote by way of network manipulation.

Caracas | 27 November 2005 | Going beyond the figures, the manipulation of 70% of the voting machines and a sleight of hand that appears to have converted the opposition’s 5 point triumph into a 20 point victory for the Government during the recall referendum, what can be deduced from the report prepared by an interdisciplinary academic team made up of 30 Venezuelan professionals is how technology has been turned into a political instrument capable of undermining the democratic system with hardly anyone noticing it.

With their names remaining confidential, three of them spoke with El Universal about the discovery of a "perfectly exportable [system] that allows a leader to perpetuate his stay in power."

—What does your team believe was the true result on the 15th of August?

—We simulated the flow of voters in the manner of a production line, calculating the times, and we found that for 8.5 million Venezuelans to vote, the official figure furnished by the CNE [National Electoral Council], 22 hours would be required, allowing 30 seconds for each one. Assuming the 19 thousand voting machines were up and running. That is six hours longer that the voting process lasted. We figure, according to the proportions, that between one and a half and two million virtual votes were inserted. These votes were attributed to voters who did not go to vote. For us this would mean that the Yes-vote must have won with a lead of between five and ten points, which they (CNE) turned into a victory of almost 20 points.

—You and your team say that you can demonstrate that there were anomalies in the flow of information, but…

—Anomalies linked to the number of voters per table, which contradicts the offer made by the CNE concerning the method to be used. The conclusions are clear: traffic prior to the official closing of the event; bidirectional data transmission in unexpected volumes; the presence of unexpected data transmission patterns that allows one to infer the existence of more than one version of the voting and scrutiny system incorporated in the machine, in contradiction to the guarantees offered by the CNE, and more than 70% of the machines showing unexpected behavior. That is to say, they were manipulated. To this one must add that 70% of the returns were not reconciled with the registration books.

—If between one and a half and two million votes were inserted, how were they able to account for the fingerprints and signatures of those presumed virtual voters?

—That question has its answer in the tangible electoral books, which we requested to the CNE, specifically to the Information Bureau, and the answer we were given was that those books are not available for examination and that their disclosure was prohibited by order of the National Electoral Board and by the Plan República [a military plan designed to monitor elections], who are the custodians of this information. Thus, since the books are not available, and after noticing the discretionary manipulation of 70% of the votes, we can conclude, based upon technical criteria, that this electoral process was adulterated.

—That is say, that there was fraud.

—Exactly. Had we had access to the electoral books, we would have been able to verify the authenticity of our conclusion, according to which this was not an honest election.

—Is it feasible to use this same procedure for the 4th of December elections, taking into account the fact that modifications and concessions have been made by the CNE?

—We do not know what directives have been issued by the CNE, because its decisions must be published in a directive and, as of this moment, nobody knows of one. Such is the case that having a connection between the machines and the server, before printing takes place, continues to be the practice. During the recall referendum the fact that there was an upstream connection was important and it gave the CNE time to carry out the manipulation. We have no idea if this time it will be different or if the machines are equipped with a wireless device. These machines, originally meant to be lottery terminals, have not been certified. It follows that not all of them are alike and therefore one cannot speak of a random sample. What happens if I open up a machine and during the audit it does not coincide as to the number of votes? The error can repeat itself in the closed boxes. That is to say in 57% of the machines. And that error favors whoever controls the process.

Translation by W.K.

© by & the author