Can you believe Venezuela’s pollsters?


Over the last year, several correspondents in Venezuela have repeatedly attempted to portray Venezuelan President Hugo Chavez as an unpopular leader. The most common basis for these statements has been the recitation of “polls” claiming that Chávez’s approval rating is down to around 30 percent.

The commercial media correspondents rarely cite the source of their polls. So this reporter contacted them, and most of the reporters offered only the names of two Venezuelan companies – Datanalisis and Keller and Associates.

An investigation into the operations of these two Venezuelan polling firms and their relationships with correspondents reveals that, by any fair measure, it is irresponsible for correspondents to cite the two firms’ polls without also mentioning that the two firms are headed by virulently anti-Chavez figures who frequently use polling samples that are unrepresentative of the overall Venezuelan population.

The first factor that calls the polls into question is the well-known political partisanship of the polling firms’ directors, Jose Antonio Gil Yepes of Datanalisis and Alfredo Keller of Keller and Associates.

In a recent e-mail interchange, The Los Angeles Times’ correspondent T. Christian Miller acknowledged that the two pollsters are “pretty anti-Chavez,” but he defends their credibility on grounds that “both do door to door polling, to get the poorest of poor represented in their surveys, and also balance for things like gender and region.” Miller’s defense of Keller and Gil Yepes is very questionable in view of contrary evidence. However, before presenting this contrary evidence, we would like to point out the problems with the two pollsters’ political partisanship.

Datanalisis’ Pollster: Chavez “has to be killed”

Gil Yepes and Keller are not merely “anti-Chavez”; they are openly and virulently anti-Chavez. In a July 8 article in the Los Angeles Times, Miller describes Gil Yepes as a man of “Venezuela’s elite” who “moves in circles of money, power and influence” and “was educated in top U.S. schools.”

It’s certainly shocking that the LA Times quoted Gil Yepes saying that Chavez “has to be killed.”

But it is even more shocking that the LA Times and other commercial media continued to use Gil Yepes’ polling “results” after his homicidal fantasies leaped out of the closet through the pages of last July’s LA Times.

According to T. Christian Miller of the LA Times, Gil Yepes saw an assassination as the only way out of the “political crisis surrounding President Hugo Chavez.” Gil Yepes has since claimed that his quote was taken out of context, and that he was only making reference to an oft-expressed sentiment among Chavez’s opposition.

But let’s look at the full context as reported by the LA Times:

Jose Antonio Gil is among Venezuela’s elite.

He moves in circles of money, power and influence. He was educated in top U.S. schools. He heads of one of the country’s most prestigious polling firms.

And he can see only one way out of the political crisis surrounding President Hugo Chavez.

“He has to be killed,” he said, using his finger to stab the table in his office far above this capital’s filthy streets. “He has to be killed.”

One need look no further than Datanalisis’ website to find the kind of blatant political partisanship that one normally does not associate with respectable polling operations. For example, in Datanalisis’ summary of a July 2002 report, the polling firm absurdly characterizes the current political conflict as one between the government (“el oficialismo”) and “the rest of the country.”

Despite the preposterousness of this portrayal, it is nevertheless an appropriate demonstration of the deep-seated class hatred by a large segment of Venezuela’s business-led opposition, which prefers to pretend that thousands of poor and working-class Chavez supporters do not exist.

When a massive pro-government demonstration in Caracas on October 13 showed that a good portion of “the rest of the country” supported Chavez, the editorial board of Venezuela’s elite-controlled newspaper El Nacional was incensed. El Nacional, which commissions and publishes polls by Datanalisis, disparagingly referred to Chavez’s supporters as “lumpen” who were lured from the country’s interior with “a piece of bread and some rum” to “come and cheer the great con man of the nation.”

As the Venezuelan anthropologist Johnny Alarcón Puentes points out, the terms “lumpen, rabble hordes, drunks, riff-raff and mobs are only some of the epithets foisted by the wealthy on citizens of dark skin, on street merchants, on workers, on the indigenous and on all those who live in slums or modest neighborhoods and dare raise their voice against the powerful.”

Thus, from the warped perspective of much of the opposition, Datanalisis’ contention that “the rest of the country” opposes Chavez makes sense. Since elites are the people that “matter,” and those of less privilege can be reduced to virtual sub-human status, poor and working-class Chavez supporters do not qualify as part of “the rest of the country.”

Alfredo Keller’s “Fight to the Death”

As with Gil Yepes, there is good reason to believe that the pollster Alfredo Keller has come to advocate a violent solution to Venezuela’s current political conflict. In Keller’s recent letter published by PetroleumWorld.com, he describes the current political standoff as “a fight to the death for power between two counter-posed ideological forces: an authoritarian socialism with a spirit of revenge against a democracy that is open to the market.”

The charge of authoritarianism against Chavez is weak, and is especially hypocritical coming from the likes of Keller.

Here is a country, wracked by unrest, provocation, sabotage and calls for political assassination, a country that suffered a 48-hour military coup last April, where the television media and commercial dailies routinely exhort the public to violence, but the Chávez administration has not arrested or imprisoned a single journalist or opposition leader.

In fact, Chavez often comes under friendly criticism from the left for being too soft on his opposition. Cuban President Fidel Castro recently remarked, “If I have something to regret, it’s his excessive generosity and kindness.” Castro continued:

“In what country could there be a coup and then have all the perpetrators meet in a plaza to spend 50 days agitating through television networks, proposing another coup? Not in any country in the world. I believe that there is not a more democratic, more law abiding, more tolerant, more generous man than Hugo Chavez.”

The authoritarian label is more applicable to Keller than to Chavez. After anti-Chavez Generals led a short-lived coup d’etat against the Venezuelan President and turned over power to businessman Pedro Carmona and his entourage of right-wing ministers, Keller called the coup “a de facto referendum”. As Carmona announced the dissolution of Venezuela’s democratically-ratified constitution and democratically-elected congress, Keller peddled the lie that the April 11 opposition march on the Miraflores Presidential Palace had forced Chavez to resign.

Evidence that emerged later suggests that opposition Generals coordinated the shootings of protesters on April 11, with the objective of using the killings as a pretext to depose Chavez and claim that they had rebelled against his supposed orders to open fire on the people. A CNN video photographer, Otto Neustald has admitted that, two hours before any killings had taken place, he filmed a rehearsed press statement by the anti-Chávez Vice-Admiral Héctor Ramírez Perez that Chávez was “massacring innocent people with snipers.” While Neustald makes clear that Generals allied to the opposition had foreknowledge that snipers would be utilized, the U.S. and British press corps in Venezuela has maintained a blackout of Neustald’s admission.

The real concern for Keller and his avaricious cohorts in the opposition is the “structure of power” that Chavez and his supporters have erected. Steve Ellner, a historian who lives in Venezuela and specializes in the country’s labor movement, has pointed out that Chavez’s reforms, which include agrarian reform and severance benefits for workers, “have strongly favored labor at the expense of business.” Some of these reforms are enshrined in the country’s new constitution, which was democratically ratified by the electorate in 2000. The majority of political representatives in the country’s new unicameral congress support the reforms.

In his recent letter, Keller expresses fear of the possibility that Chavez could still be in power by August, the month when the constitution allows for a binding referendum on the fate of the government. Although Keller claims that Chavez would lose such a referendum, he says that a political transition of that sort would still represent “a tremendous defeat for the opposition” because the “structure of power… would remain intact.”

Like coup leader Carmona, zealous figures within the opposition such as Keller seek to erase the entire Chavez legacy. But since that legacy has unleashed popular social forces that will rightly resist a return to oligarchic rule, the insistence of Keller and other opposition figures’ on such an uncompromising position suggests their willingness to promote violence.

Partisan Pollsters

The known political partisanship of Venezuela’s pollsters causes all sorts of problems with regard to their polling. Firstly, it calls into question whether or not they are posing survey questions in a non-biased fashion. But as any political consultant will admit, a pollster, by phrasing the questions and deciding the “survey sample” of how the poll is “weighted” to specific demographic groups, can get any result he wants.

But even if we were to assume that Keller and Gil Yepes are not loading their questions, the poll respondents’ simple awareness of the pollsters’ political partisanship is likely to skew the polls in favor of the opposition.

We asked Matthew Mendelsohn, a Canadian political scientist and specialist on polling methodology, whether or not the pollsters’ well-known political partisanship — independent of all other factors — could bias polling results. Although Mendelsohn told us that he lacked knowledge about polling in Latin America, he responded as follows:

“Any perception on the part of the respondent that the questioner is partisan can influence results. You see this with interviewer effects all the time — male and female, black and white, etc. interviewers get different results. And certainly if the respondent knows that you’re a representative from a particular party or group, this biases results.”

Biased Polling Samples

The factors that are likely to bias the polling of Gil Yepes and Keller are not limited to political partisanship alone.

An academic source — a person that has worked closely with Venezuela’s pollsters – said that most of Keller’s polling has been done in the middle class areas of the ten largest cities, meaning that the populous slums where Chavez’s support is concentrated have been largely excluded from Keller’s polling sample.

Our source informs us that Datanalisis’ polling samples are less skewed than Keller’s due to the firm’s superior operational team of field workers and access to Venezuela’s 1998 census tracts. However, the poll that Gil Yepes is currently releasing about the population’s views of the so-called “general strike” and Chavez’s handling of the crisis appears to be highly deceptive.

Here’s another fact unreported by English-language correspondents who cite polls by Gil Yepes and Keller as gospel: Since the “strike” began on December 2, Chavistas are not allowing Datanalisis’ field workers into the Chavista-controlled slums of Caracas and Maracaibo. While Gil Yepes recently released lopsided polls that purport popular support for the “strike,” he fails to mention that his polling sample excludes the populous slums where the “strike” has proved to be a complete failure. The progressive economist Mark Weisbrot, who recently spent time in Caracas, wrote a column for the Washington Post explaining that there were “few signs of the strike” in “most of the city, where poor and working-class people live.”

The academic source said that Keller and Gil Yepes generally do not poll rural inhabitants. The opposition newspapers that commission the polls are not willing to pay the increased costs that rural polling entails. Thus, landless peasants who may benefit from Chavez’s agrarian reform are also excluded from polling samples.

Tainted Pollsters, Tainted Press

In view of the above-mentioned facts, it is mind-boggling to see just how laudatory the English-language press corps is of Gil Yepes and Keller.

AP’s Alexandra Olson calls Datanalisis’ “Venezuela’s most prestigious polling firm” in a recent report.

The Miami Herald’s Juan Tamayo claims, in an e-mail reply to this reporter, that Datanalisis and Keller and Associates are “the two most credible polling companies in Venezuela.”

Jehan Senaratna of Dow Jones News Wires calls Keller “the head of a respected Caracas-based polling and economic research firm.” Despite his polite remark about Keller, Senaratna tells us that Datanalisis is the “only polling firm that can be considered reliable and unbiased politically.”

Finally, Phil Gunson, a freelance correspondent in Venezuela who has written for several papers, says the “polling organizations that most of us consider to be the most reliable” are Keller and Associates and Datanalisis.

In essence, the correspondents have become so carried away with anti-Chavez hysteria that they are blinded to the fact that the pollsters whom they rely upon are neither credible, reliable, or politically unbiased. How would Keller and Gil Yepes be received in other lands, even in the United States, promoting themselves as respected pollsters while making statements that verge on inciting violence against a democratically elected government?

So the next time a member of the commercial press corps tells you that umpteen percent of the Venezuelan people feel a certain way according to “polls,” ask yourself: Did they identify the source of the “poll”? And if the “poll” was about Venezuelan President Hugo Chávez, was the pollster someone in “a fight to the death” or who says “Chávez must be killed”?

In a media-fed democracy, polls and simulated polls can be lethal weapons, too.

Justin Delacour is a freelance writer and recent graduate of the Masters program in Latin American Studies at the University of New Mexico. He has written for Latin America Data Base (http://ladb.unm.edu/), a University of New Mexico-based news service. He receives email at [email protected]

Leave a comment