Polls from hell, and Reuters taking them at face value

Long time readers know that I have long ago stopped putting major credence on Venezuelan polling companies.  I only follow Datanalisis as relatively more accurate than others (it did good in the recent primaries for example) and Keller who in spite of all his troubles may be the best one at illustrating tendencies.  Consultores 21 is not bad too.  But no matter what, in the country of the Tascon List, it is wise not to rely too much on polls and more on your own historical analysis since too many people are unwilling to say what they really think.  Such skepticism has served me well, allowing me to predict the result of the legislative elections and the one of the primaries, and a rather good score on the 2008 regional ones.

This being said, it is necessary to break my rule and comment the latest of Datanalisis, that we know only through Reuters, written by Marianna Parraga who has more than once stumbled upon her Venezuela reporting with booh-boos that cannot be easily forgotten, to the point that many suspect her to be pro Chavez.  Far from me to go deep into her latest work because, well, it is a shoddy  piece of journalism, relying on a single undisclosed source for a private poll of which we know no detail, etc, etc...  Blogs, such as yours truly, are written with more seriousness than her Reuters reports!


No, what I want to address here is that so many people, in spite of all the evidence, are only too happy to buy such polls that predict a victory for Chavez even bigger than in 2006.  My point, for those still with me so far, is to remind a few truisms that should not be forgotten.In 2006 the result was Chavez 7.3 million (63%) and Rosales 4.3 million (37%).  That is, a victory by 26 points.  Today we are told to assume that Chavez is going to win by at least 20 points. I simply do not buy it, not even from Datanalisis, the more so that to this date we have no details on their poll where Capriles is behind by 16 points.  Why can I say that?

First, from my analysis of the primary result I am convinced that the electoral starting ground for Capriles is 5.3 million.  Unless his campaign were to collapse, which it is not doing yet, he cannot get less than 5.3 million votes.

Second, the last term of Chavez has been so calamitous that he simply cannot get more votes, EVEN if the Capriles campaign were to collapse.  Chavez will not get more than 7 million votes (assuming no cheating of course, but that is another story).  That is, in his best scenario Chavez wins by 1.3 million votes, or 57 to 43 points, 14 points maximum spread!

Third, not only Capriles launch was a resounding success, but his posterior events all around the country have been a big hit considering that his campaign does not have the means to ferry people the way chavismo does, requisitioning buses and forcing public employees to attend.  I understand that after 14 years of chavismo the enthusiasm is naturally in the opposition side as the chavista camp is voting more out of duty or habit than anything else; but still, such enthusiasm is not the mark of a collapsing campaign even though I am among the ones who think that the campaign is not the one we would wish to see.
In other words, I see nothing that would make me reconsider my calculations of last February, not even the way Chavez has skilfully managed his cancer as an emotional blackmail to his electorate.

The way I see things now, today, at present, is an even race with Chavez, still leading by a default of sorts of the electorate which is not used to the idea that things could change. But the official campaign has not started yet and even if today we may really be 47 to 53 in favor of Chavez (I cannot see it any higher than that today), there is still plenty of time for Capriles to turn around these numbers by election day.

If you really live and die by pollsters I would advise that you wait for a real Datanalisis poll (and not the one for a "private" client that shall remain nameless), the Keller trend that should be out by August and Consultores which has been silent long enough that soon we should hear from them. Make an average and live by that number. Be aware that chavismo has launched itself into a through and through polling war, with paid pollsters that used to be considered serious before (any one reading from Hinterlaces?). Surely such polls with the extended coverage in state media have an effect, albeit temporary.

Or take a deep breath, relax and come back here and read your favorite blogger number crunching that should start sometime in August :-)