TCS Daily

Venezuela Recall Update

By Carroll Andrew - August 23, 2004 12:00 AM

At the time of this writing, the leadership of the opposition to Venezuelan President Hugo Chavez, the Coordinadora Democratica (CD), has not officially accepted Chavez's 58% - 42% victory in the August 15 recall referendum. On Friday afternoon, CD leaders presented evidence to Organization of American States (OAS) and Carter Center election observers alleging electronic tampering with the balloting. Chavez's government, unsurprisingly, dismisses this charge and demands that the opposition accept its defeat. The OAS and Carter Center, having observed a government audit of the election results, report no evidence of fraud.

Venezuela's opposition has, to date, satisfied the most important rule of skeptical inquiry; they have provided verifiable claims. Their delay in accepting the result is, therefore, reasonable. Enrique ter Horst summarized the basis of the fraud claim in an August 18 International Herald Tribune article,

"Three machines in a voting center in the state of Bolivar that has generally voted against Chávez all showed the same 133 votes for the Yes option, and higher numbers for the No option. Two other machines registered 126 Yes votes and much higher votes for the No. The opposition alleges that these machines, which can both send and receive information, were reprogrammed to start adjudicating all votes to the No option after a given number of Yes votes has been registered." [emphasis mine]

Media reports from individual Venezuelan states supported investigation of the fraud claim. El Universal reported claims that 25% of the data from Aragua state and 52% of the data from Bolivar state had shown repeated "Yes" total irregularities. Venezuela's Union Radio website reported that 33% of the polling places in Zulia state involved irregularities.

On Friday morning, the Carter Center endorsed a set of numbers contradicting claims of widespread tampering. They analyzed data from the 8,000 "mesas" (literally, "tables"; in this context, polling sub-stations) involved in the election. They found 402 mesas containing at least two machines with identical "Yes" totals, and 311 mesas containing at least two machines with identical "No" totals. They claimed, as a percentage of the total of 8,000, the number of identical results was statistically insignificant.

On Friday afternoon, another set of numbers came into play. The CD presented OAS Secretary General Cesar Gaviria with 1,879 "actas" -- printouts of totals from individual voting machines. (As part of Venezuelan electoral procedure, the opposition has the right to have witnesses obtain copies of final voting machine tallies). Each acta was from a voting machine that had registered a total identical to another machine in the same polling station. In addition, the CD produced evidence relating to a set of so-called "hot audits" -- audits conducted on the day of the referendum -- showing "Yes" winning by a comfortable margin before the capping effects would have begun.

An analysis of machine-by-machine totals that became available by the end of the week showed that the Carter Center and the CD were probably presenting the same results in different ways. The Carter Center's total of 402 was a count of the number of mesas containing two or more machines with identical totals. The CD counted the data by polling station, not mesa. Each polling station included 1 to 3 mesas, increasing the number of opportunities for totals to repeat themselves. In terms of polling stations, there are 805 cases where the same vote total was repeated on two or more machines. The figure of 1,879 is, apparently, the number of voting machines involved in these 805 stations.

Now that the machine-by-machine vote totals are available, experienced and impartial electoral analysts, scattered across statistics and political science departments on every continent, can have at it. They will be able to determine if the results contain any conclusive evidence of tampering.

Who Audits the Auditors?

On Thursday evening, the CNE (the body that oversees Venezuelan elections) began an audit of the election results. The process was observed by Carter Center and OAS representatives. On the day of the election, immediately after voting, each voter had received a paper receipt indicating his or her choice. The voter placed that receipt into a ballot box. The CNE compared the official, electronic tally to paper ballots produced at 150 different polling stations. On Saturday afternoon, the results of the audit of the paper ballots were used to affirm the legitimacy of the election result.

The CD, dissatisfied with the audit procedures, had refused to participate in this audit. They did not want to endorse a process that they felt was insufficiently sensitive to the possibility of electronic fraud. The CD fears that it is possible that the paper ballots reflected the voters' actual choices, that the electronic results were altered in selected locations, and that the CNE has the means to work around locations where fraud was committed.

A more complete audit of the result would help allay these fears. Much of the criticism of the audit centers on the fact that the CNE -- supposedly randomly but definitely unilaterally -- chose the 150 polling stations without input from the CD. Is it beyond imagination to investigate the possibility that the CNE worked around stations where they knew that fraud occurred? Were the centers to be audited chosen well enough in advance so that there was time to manufacture fake ballots?

Why not let the CD identify the stations arousing their highest suspicion and let them be audited? Antonio Mugica, the president of Smartmatic, the company that supplied Venezuela's voting machines, has said in El Universal that he is committed to "all the audits that are necessary." Given the stakes, it is reasonable to ask Mr. Mugica to honor his commitment.

The author is a frequent TCS contributor.


TCS Daily Archives