ELECTION 2013 | Interview with pollster Afredo Weil

"Capriles won by 52%, with a four-point slide"

"In Zulia (west Venezuela), guerrillas controlled a polling station (...) insofar as null votes increased, voting in favor of Capriles diminished"

In Weil’s judgment, the shadow of fraud has always been present (ADOLFO ACOSTA)
Monday April 29, 2013  05:37 AM
Alfredo Weil does not tell a number, or concludes an idea, or makes a statement, without consulting Wilson, the Ipad called in this way by his wife, where he stores a huge file of data and statistics. Based on this, the director of pollster Esdata upholds that all polls, from 2004 to date, have been marred by the shadow of fraud and that opposition candidate Henrique Capriles Radonski won the election of April 14 with a four-point leeway.

"If we take official data from the CNE (National Electoral Council) into consideration and assess balloting places with three or more polling stations, we find that Capriles won by one and a half million votes. Furthermore, if you take balloting places with two polling stations, he lost by 332,000 votes. Nevertheless, if you pool these two factors, plus votes abroad (50,000), Capriles takes the lead by 263,000. Now, when you count balloting places with one polling station only (over 5,000 or 8.86%) Maduro wins by 477,000. As for the rest of the country (91.14%), the victory goes to Capriles."

Do you refer to balloting places with one polling station which are generally located on remote sites in Venezuela?

This is what you think, but this is not the case. It happens as well in Caracas parishes, such as Coche or Santa Rosalía.

Why does it happen in places where witnesses are supposed to be and oversight is more active than in a rural environment?

The fact of the matter is that witnesses' control is very difficult and they are precluded from doing their work. In Zulia, for instance, at José Ramón Yépez Parish, Maduro wins 57 vs. 42. However, when you inquire into what happened to Bolivarian schools, where balloting places operate, you will find that Capriles won in some and lost in some others. There are balloting places, as this one, where the difference is 99%. Thus, with such results, you "kill" the entire parish.

Is there an overall trend?


Could you define it?

They are strategically located places, where people of an overwhelming red (pro-government) trend are enrolled or mobilized. Therefore, finding a witness (for the opposition Unified Democratic Panel to vouch for the results) is impossible.

Could you substantiate such contention?

The information available to us on this balloting place, in Zulia, is that Che Guevara Elementary School is controlled by guerrillas. Not even Plan Republic (the deployment of military troops to safeguard balloting centers and voters) enters there.

In how many balloting centers did it occur?

They are more than 3,000 balloting places throughout the country, and this was also the case on October 7, except for a more effective government organization. Despite the death of (President Hugo) Chávez and the mourning, the leeway for the government downsized 1,300,000 votes by matching the results of April 14 with those of October 7.

Did any other factor influence voting?

Null votes, which record a numerical curiosity; insofar as they go up, voting Capriles is adversely affected.

Do you mean that those votes were for Henrique Capriles Radonski?

They were for Capriles and appear like null and void. It is just that they do it in a commensurate way. Thus, a null vote per polling station may result in 40,000 nationwide. However, if you take three null votes per every polling station, the number will advance to 120,000 votes. Add to this the votes missing during the transfer from the balloting place in Miami to New Orleans, the closure of borders, "assisted" votes, coercion, blackmail and other irregularities. You could reach a number that, based on such a narrow result, could make the difference.

Capriles has asserted that the votes for him, recorded by the CNE are what they are. This means that actual voting for (Venezuelan President Nicolás) Maduro was below 7,200,000 votes.

There was a universe of 18,903,000 enrolled voters; you have 283,000 null votes, and, most importantly, almost 80% turnout. Nevertheless, such is a bloated number. You can match turnout levels in any country in the Western Hemisphere and will find that abstention is around 30-35%. This is what should be recorded in Venezuela.

This was the case on October 7.

On October 7, Capriles got 6,591,000 votes, whereas Chávez hit 8,191,000. But he died and there was mourning. In light of this event, election was held on April 14. What did we find? Despite the CNE's denial, the register of voters was expanded by 50,000 voters, with somewhat the same turnout. I think, though, that a fervent Chávez's follower benefitted from government policies, compelled to vote a candidate like Maduro, would hesitate. And this might influence abstention. As for me, numbers do not tally so that Chávez's followers would swap their votes in the proportion recorded on behalf of Capriles (700,000 votes), particularly if there is no abstention.

Chávez's followers did not stop voting.

Exactly, and this is precisely what is hard to believe. There must be a percentage that could reach, as it were, 3% that would say: "I will not vote that guy."

Nevertheless, Capriles gained 700,000 votes with the same abstention percentage as in October.

Chávez got around 8,200,000 (round numbers), and Maduro got 7,570,000; about 600,000 fewer votes...

And Capriles amassed...

More than that.

However, the number of voters lost by Maduro and the numbers of votes gained by Capriles are much alike.

If I am pro-Obama and he trips, makes a mistake or tells something politically incorrect, I will opt not to vote Obama. This does not mean, though, that I will vote Romney. I would tend to abstention.

Therefore, why Capriles got such an amount of votes?

Because this unveils, based on better control, those votes really existed on October 7; now, this comes to the fore. And when this occurs and they (the government) realize that you are recovering the actual value of your electoral support they enter a stage of alarm. "What can we do to overtake such number?"

Are you inferring that there was also fraud on October 7?

As far as I am concerned, all ballots, including those of 2004, no matter the opposition winning or not, show a factor of implicit fraud.

Why did they fail to allocate Maduro those 700,000 votes?

J.J Rendón (a Venezuelan image advisor and expert in political communication) said that in order to overcome fraud, the opposition should win 70/30. I am afraid that Capriles, in a clean election, would not win by such a leeway, yet it is possible for him to strike 60/40. Therefore, you should cover 20 points, that is, three million votes, a number too hard to digest. However, they attempt at reducing it.

Which way?

There are inconsistent results in polling stations located in the same balloting place. That is, it does not make sense, from the statistical view, to have a polling station where Capriles won 20/40, but in the polling station right beside it quite the opposite occurred. The probability of this to happen is measured by the Pearson's test. And this test allocated a 0.1% probability to such a case. Thus, insofar as the probability of refusal was higher in polling stations, Capriles' votes lowered.

On October 7, we took a sample of polling stations with a refusal probability of 10% or more. And Capriles could win by 51% or lose by 48% down to 44%.

What did happen on April 14?

If we get rid of everything over 10%, Capriles could have won by 56%. Insofar as the probability of refusal rose, Capriles would go down. However, with 60% and even 90% of refusal, Capriles would win anyhow. Therefore, the most plausible result is 52/48. More accurately: Capriles 51.9, Maduro 47.8. This is our result based on our mathematical and forensic analysis. It does not mean absolute truth, yet an indication.

On which grounds did you premise to reach such conclusion?

Let me explain with an example. Baruta is a par excellence middle-class municipality. There, in Santa Rosa de Lima, Capriles won 67/32. The CNE allowed opening half of the ballot boxes, which would be an acceptable sample. However, in default of trick, one single ballot box would suffice. Now, go to the low-income area of Minas de Baruta municipality. You will find notorious differences from one to another polling station located in the same balloting place. For this reason, we ask for recount of votes of all ballot boxes.

Translated by Conchita Delgado
The end of a cycle

Hundreds of thousands of demonstrators took to the streets of Brazil on March 13 to demand the ouster of embattled President Dilma Rousseff, carrying banners expressing anger at bribery scandals and economic woes. A banner read "We don't want a new Venezuela in Brazil."