Et tu, CAF?

Chávez Always Forever

Suppose you were the head of a multilateral bank, organizing a forum dealing with a crucial, upcoming election.

When discussing what the polls are saying, do you ask someone who’s already made up his mind about who’s going to win to give you a speech? Or do you pick someone impartial, above the fray, who can lay out the different scenarios for you?

Seems like an obvious question, right?

Turns out the Corporación Andina de Fomento is holding its annual meeting in Washington this week.

CAF is a venerable, old multilateral lender, usually unmarred by political maneuverings and focused on making loans for development projects all over the continent. It may be based in Caracas, but this is not Banco del Sur we’re talking about.

So who did they invite to talk about the trends in public opinion polling a month before the election? None other than our friend Luis Vicente León, the visible head of Caracas polling firm Datanálisis.

CAF knows that there is currently a polling war going on in Venezuela. On one side, we have a group of firms led by Datanálisis that basically say Capriles has no chance. On the other side, we have a group of firms headed by Consultores 21 that say both candidates are tied, and that Capriles has momentum.

Instead of choosing someone balanced to give a talk about what might happen, CAF has decided all the attendees at its conference will hear only one side of the story.

CAF is basically conceding the election to Chávez at a moment when things are quite fluid. That is a mighty tricky – and intrusive – position for a multilateral bank to assume, particularly concerning the election for the country where its headquarters are located.

These days, chavismo is intent on creating the opinion that Chávez is leading comfortably, and that the opposition will not accept a win by Chávez. It wants to create the opinion that the election is over, that the result is a foregone conclusion.

And CAF is playing right into their hands.

Now, León may be right and Chávez may win comfortably. But if Capriles manages a win, this will shatter Datanálisis’s reputation forever. This will also reflect poorly on those, such as CAF, who have given León a forum and selected him as the only legitimate pollster in Venezuela.

CAF’s gamble is a risky one. Could it be due to the fact that they’ve done well under Chávez? Regardless, we’ll be sure to hold their feet to the fire in case Capriles manages a win.

Caracas Chronicles is 100% reader-supported. Support independent Venezuelan journalism by making a donation.


  1. Datanalisis is not saying crazy things. They place Chavez with only 43% intention of vote. Capriles has 27%. The rest is NS/NC. Luis Vicente Leon knows that Chavez is not that popular. He must know that the amount of NS/NC is due too shortcomings in his interviews. If a little more of those who are in the category of NS/NC prefer Capriles then Chavez is screwed. And my guess is that if they are not answering is that they fear to voice their opinion, and not so much that they don’t know.

    Historically those NS/NC has transformed into votes to the opposition while abstention at a national level usually hurts Chavez (former chavistas prefer to not vote that vote against Chavez).

    • No no no, let’s call a spade a spade. Datanalisis believes Capriles has no chance. Consultores 21 is saying they are tied. They are both good companies, but obviously one will be proven wrong and the other one won’t be. CAF has decided to side with one of them – curiously, the one that gives their preferred candidate the edge.

      (and don’t get me started on CAF’s complicated relationship with Capriles and his administration… they’ve had a strained relationship for years now, which may explain their choice of speakers)

      • “Datanalisis believes Capriles has no chance.”

        My problem here is with the word “believes”.

        The point isn’t that LVL “believes” this or that, the point is about what the data Datanalisis is collecting shows.

        Do you have any sort of evidence that Datanalisis’s data collection methods are corrupt? That LVL’s beliefs are not grounded in his data? And need I remind you that C21’s polling has been more systematically biased in previous elections than Datanalisis’?

        There’s a strong whiff of Shoot-the-Messenger to your stance on this, Juan, and I think it’s sort of distasteful. Maybe LVL is wrong, and if that’s how it turns out you’re extra-welcome to make all kinds of fun of him on 8O. But when you teeter from saying he’s wrong to suggesting he’s corrupt, that’s where I get off the bus. We just have no evidence of that…

        • YV Polis is picking and choosing his data points.

          And no, I do not believe Datanálisis is biased, nor do I believe C21 is biased. But they can’t both be right.

          And when did I say LVL or Datanálisis were corrupt? Don’t put word in my mouth.

          • Maybe CAF should not invite to only one source. But I am not sure that your claim about YV Polis are correct. Juan can you explain in which way YV Polis is picking and choosing his data points? I think he ihas used data from several elections and data from all type of “encuestadoras” pro-gov. and pro-oppo.

          • He is not using data points from regional elections. I think the accuracy of polling firms in recent regional elections says just as much about their methods as their accuracy in a presidential election six years ago.

          • I see my name here and I have to jump in…Juan, the only reason I am not including regional elections is because I don’t have data (I think I mentioned this before). If you get me the polls I’ll add them to the analysis.
            In any case I think your argument at the beginning is right and you should start a petition to get ME invited to give that speech =-).

        • The whole pollsters war makes my head spin. Didn’t someone say here that the data from Datanalisis comes from a much bigger study where they poll a lot of things and by the time people get to the questions about vote intention they are giving up in droves and that would explain the high number of NS/NC? I seem to remember that from a older comment and thought the conclusion was that we needed to see other polls from Datanalisis that were specifically about the election to see if they would find the same NS/NC…

          • Unfortch, I tried to verify the Questionnaire Length hypothesis but it doesn’t fly – C21 and Datanalisis questions come at the same point in their questionnaire. They’re also identically worded! The only thing that varies between them is the polling points, but that really doesn’t explain the gap. Both firms are aware that something WEIRD is happening here, and I don’t think either is happy about it.

  2. Dear Juan:

    Your message is totally biased and misleading. CAF is not suggesting anything. CAF has its own right to invite any pollster to its annual event and that does not mean CAF stands any comment or conclusion about a specific poll result. I did not understand your point. If you invite a friend to your house then automatically you share his/her points of view? It does not make any sense. If you go to Washington or any other place and listening a pollster opinion then you are able to disagree or agree with that, period. Final words: C21 and Datanalisis are the best public opinion firms in Venezuela. Historically they have published similar results until this election. Is it mean that Datanalisis is lying or intentionally biased? I do not think so. At the end, what you´re claiming (impartiality) is what you did not do in your article. In any case, if Capriles wins, I do not have any doubt that CAF will support his development program because it is its duty. Same applies to a Chavez victory.

    • Dear Alfonso.

      I have never made any claim to be impartial. CAF, however, does. By wading into this debate and providing only one point of view, however valid it may be, it is wading into a hot political debate and indirectly helping important people form an opinion, one which – I acknowledge – may be true in the end.

      Sadly, most of the people going to the conference in Washington will come away thinking Datanálisis’ numbers are the be-all and end-all. Why, they come with CAF’s seal of approval!

      Sorry, CAF made a blunder, and it’s interfering in my country’s election. I have a right to voice my opinion.

      • Juan, you are right in this comment. Washington is a key forum for U.S.opinion, and CAF is just protecting their comfortable (supposedly profitable) relationship with the terribly corrupt Chavez government. And, no, I do not believe in Datanalisis’ purity. As I’ve said here before, I was present at a private 150 or so high corporate meeting in Caracas 2 months before the large Chavez 1998 Presidential electoral win, and Gil Yepez himself scoffed at even the remotest possibility of a Chavez win. So, either he’s corrupt, or he’s incompetent–take your choice–probably a little bit of each. And, no, Capriles’ winning will not tarnish him forever, as neither did Chavez’s winning–he will simply follow the money/business–Estamos en Venezuela, compadre, remember?? Finally, recently a key Chavista street level organizer, who has been doing that for 14 years, especially in the Interior, told me that Capriles/Co. are going to win ALL of the governorships–so, HOW is Chavez going to win the Presidency–because he’s been so effective at helping this POOR pueblo with his AMOR????

      • One thing, however, Juan, is that none, or at most very few of those people in Washington will have a direct impact on said election, that is for Venezuelan voters to decide… a conference about the upcoming election might help businesspeople form a view about a post-Chavez Venezuela, but I do not think that after 13 years in power, anybody wanting to do business in Venezuela is unaware of the particularities of the current administration. So what kind of influence can LVL have vis a vis the election? In the end, if Capriles wins, said businesspeople will simply adjust their business models to work with the new guy, as the always have.

  3. CAF probably is just playing it safe: “If Chávez wins, this move will be very convenient, if HCR wins, he will not take it upon CAF for having invited Luis Vicente León”. As simple as the weakly dominant strategy concept indicates.

  4. This article is a mirror version of those the Chavistas write about the US Embassy. I thought you guys were supportive of free speech? Besides which, have you thought that Datanalisis could be right? And to top it all off, the meeting is in Washington – what kind of signal does that send!!

  5. OK guys, I will ask you something, and reply seriously. Is it possible that Radonsky had 34% last July and now, out of the sudden, not only did he stop his growing trend, but actually dropped 7 points to a March 2012 start-point? Sorry, but I don’t buy it. Either Chavez or LVL or both of them are manipulating, but it cannot be right. And BTW, very sospicious that cara de queso comes out with such a survey when a ACN’s leak shows 37% to Capriles and 44% to Chavez (which is, btw, the same % he had in that leak).

    So, has anything happened these last 2 weeks that has made Capriles dropped? A bridge falls down, Amuay explodes, and so many other bad things to Chavez happen and Capriles dropped?

    NO way I will buy that shit!

    Here is the ACN’s leak

    The other thing is that Varianzas, Consultores 21, Hercon, and others are showing a Capriles 4-point lead, and then you will tell me that LVL with his 30% DK/NAs is the papá de los helados here and the others are all wrong (remember that cara de queso said 45-50 deputies for opp. in 2010).

    So, Datanalisis’ credibility died today for me! It’s not a killing-the-bird syndrome, it is rather a “I don’t believe that a million flies are wrong just to say cara de queso is right, moreover when 30% of his respondients are “undecided”!


    • Datanálisis may be right, I grant them that. But they may be terribly, tragically, horrifyingly wrong as well. We’ll soon find out.

      • Let’s see what Datanalisis says in its last poll in a couple of weeks. So far, they have reported around 20% of NS/NC. That’s huge! If in their next poll they still have a 2-digit percentage of NS/NC with a 1-digit percentage points difference between the candidates, they are covering their asses. Whoever wins, they can come out and say that the massive number of NS/NC decided the election, or not! So, basically they are saying they don’t now shit. Either their methodology is flawed because they can’t identify preferences or something fishy is going on. Either way, their reputation is probably going to suffer a blow in my book.

  6. If you check last Datanalisis poll of 16/07 through 09/08, they say the gap is (12.5) 46.8 to 34.3. Then they proceed to present several scenarios. The first talks about voters sure to vote. Definitely decided for one candidate. There the gap goes to 13.6 (39.7 to 26.1). Then they say that the undecided decreased from 23.1 to 18.8 from their last poll. And that those previously undecided that are not undecided any longer, distributed themselves 81.4% to 18.6% in favor of Capriles. Then they present a slide where they assume that only another 4 some percent of those undecided are going to be decided at the time of the election, saying that the gap reduced to 9.8%. They then say that the gap may be between 9.8% to 13.6%.

    Here is my problem with this. If those undecided are going to vote, they will decide in one way or another. We won’t get 14% of null votes in a presidential election. If you distribute the 18.8% in the 81.4-18.6 way the last group of undecided distributed, the gap ends in ONLY 0.69%. 50.3% Chavez, 49.6 Capriles.

  7. I can tell you LVL is not a chavista, nor are the people at CAF. Unfortunately, latest poll numbers remail at 60-40 for the Government (undecided voters and ninis do vote after all). I do not know what the deal is with Consultores 21 this time, but they are way off, sad

  8. 60-40 is not the scenario. All the serious analysts and what people in both sides know is that this election is going to be a toss-up.

  9. Según un twit de Ana Julia Jatar, pudo enterarse por amigos que asistieron a la Conferencia de la CAF, y la presentación de LVL fue bastante balanceada. De hecho, indicó que actualmente los números entre ambos candidatos representan un “empate técnico”. Que guabineo este hombre…


Please enter your comment!
Please enter your name here