THE POLLSTER'S WAR IN VENEZUELA: FAKE OR REAL?

Sunday 4, July 2004 (1)

What might have been the main news this week in Caracas was the publication of the partial results of the Greenberg et al. poll, that came flat out to contradict earlier polls from Datanalisis and Consultores 21. Of course this provoked some consternation within the opposition and felicity in chavismo. But was that significant? My answer is no, I am not worried at all and I will explain why at the end.

Chavez and pollsters

Like any government in trouble, the Chavez administration is discrediting any polling organization that does not publish favorable results. Par for the course, even though Datanalisis and Consultores 21 predicted very accurately Chavez victory in 1998. If we had to this that we have an administration that has long stopped accounting for its actions, that runs on emotional arguments, and a president that excoriates the local press and has not given a press conference in 2 years, one should not be surprised at the virulence presented by some notorious chavistas as to adverse polling results. (2)

The pollsters in Venezuela

There are many polling institutes in Venezuela, from the long established ones to hacks that appear at election time. There also the foreign pollsters, serious or not, that are hired on occasion. The recent example that occupies the attention of today's post is the one from Greenberg and associates from the US, that goes against the grain of local recent polls. Unfortunately, this poll complete results and methodology have not been released yet.

The polling difficulties in Venezuela

Polling in Venezuela is rather difficult. The neat split into 5 social sectors actually illustrates quite well the problem. Sectors A, B and C represent the rich, upper middle class and middle class. D and E represent workers and poor. Or so it was in the 80ies when the system was set. But now with nearly 20% official jobless rate and perhaps 50% in the informal economy these divisions are becoming meaningless. Sector A is statistically nonexistent and more and more A and B are considered as one. E, the pauper sector includes probably quite a large number of people near indigent status. In addition the E sector resides in the extensive Caracas and other cities slums and includes those that traditionally have been unwilling or unable to follow the political discourse of the country. The difficulty to poll in a meaningful way this sector should be obvious. The D sector includes more and more the fraction of the C sector that lost its economic stability but not necessarily its values and political motivations. The social changes that have happened in Venezuela in the last 10 years should bring into question some of the polling methods used in the past. Not to mention that recent political pressure could affect even further how people tend to reply to pollster questions. (3)

Recent polls

This is a little survey of recent polls and comments (4)

Pollster	Yes	No	Comments
Datanalisis	57.4	42.6	Well established Venezuelan Pollsters. Have been describing a tightening of the race but still give the greater advantage to the opposition. They regularly poll the Venezuelan situation.
Consultores 21	65.8	32.7
Mercanalisis	48-63	21-36	Asks what people think will happen, not what they will vote. Divided by regions. Also a long established poll firm.
Greenberg, Quinlan, Rosner Research	44	49	A reputed US polling firm hired on occasion. Has rather suddenly and largely reversed its general trend. Until 2 months ago it gave a large advantage to the opposition. All data and method not released yet
Datos	35	51	No web info, no methodology, except for a Venpres report
Indaga and North American Opinion research	42	55	Indaga is a recent creation from a faculty at the University of Zulia. Both are hired by the current administration. No information available on the other company though there is an extensive discussion on Venezuelanalysis, without giving information on NOAR (5)

The least that one can say on these polls, done relatively close in time, is that there is quite a variation. If the results of Indaga and Datos are to be expected, the ones from Greenberg are surprising. These are a considerable reversal of the type of numbers that they used to show. What is more worrisome is that as of today the method has not been published so one cannot decide as to the validity of these numbers. Certainly if they are right today one would wonder about the previous Greenberg surveys when the opposition had much more favorable numbers. Why such a sudden change in such a short period of time? In particular when the "traditional" Venezuelan pollsters agree on the opposition advance even if all agree that this one has shrunk some in recent weeks.

My take

Behavior of the Venezuelan electorate

As Val pointed recently, or Alek Boyd, we are victims of 50 years of populist administrations. A large fraction of the electorate is driven by what they are expecting to get, directly, from whomever is in office, or wants to get into office. This is our curse, perhaps 70% of the electorate motivated by "what is in there for me?", materialistically speaking.

Another often lobbed criticism to the Venezuelan electorate is its short term memory. Indeed, we have reelected in 1988 the president that started corruption big time in Venezuela, Carlos Andres Perez. We have reelected in 1993 the president that has lusted more for power, that has been more willing to make any deal to stay in office, one of them being the "sharing" of the judicial power, Rafael Caldera.

The arguments of the opposition

Well, they are as valid today as they were in April, even more so when one is watching with a certain impotence while the High Court is being packed, or how the oil revenue is spent without any accounting, likely in individual pockets. However there is one that might be backfiring, provisionally, for the opposition: Chavez the dictator. By accepting to go for a Recall Election in spite of all the tricks used to block the signature gathering process, a few in Venezuela wonder if the opposition was not exaggerating all along. It was not, all the other symptoms of an authoritarian wanna-be are there. But people tend to forget and that has saved more than one politician in Venezuela.

The other thing is that the opposition will have a tough job to go against some popular social programs, the ones from the "misiones". Amazingly this time Chavez has managed to strike a chord in the electorate, even if the programs do not reach as many as he would like us to believe, even if they are a den of corruption, even if they are funded at the expense of existing programs and institutions. In fact the present programs have helped erase the memory of previous failures of the social agenda such a the "Plan Bolivar 2000" who ended up enriching a few generals. Indeed, why should Juan Bimba, our Joe Sixpack, care about the packing of the High Court when he might be on the verge of getting a direct hand out? After all justice never reached him but 100 dollars would come in handy.

The opposition dilemma is to expose the cheapness and inadequacy of the present "misiones" while at the same time stating that they will strengthen them, thus indirectly validating them.

So, what are the numbers?

Well, I still think that my detailed table of a few days ago is still valid. I would just add that Chavez has been able to add, albeit temporarily, the few points that are up for grab. He benefits from the positive effect of gaining back some democratic shade, of playing nice, from the "misiones" where he is pouring money while accusing the opposition to threaten to take them away. He also benefits from the down time of the opposition, enjoying his May success and gearing up for totally new type of campaign. Chavez had started his campaign even before the signatures were in and that helps in polls. The reader will remember that I have always given Chavez at least a 30% hard core vote, and if he reaches today a 35-40 % that would give him, a slight edge over the opposition once abstention is factored in.

But this week end is the official start of the campaign. Let's see what happens in the polls after 5 weeks of reminding people the failures of Chavez and why they had to go and sign up. Let's see what happens when Chavez blows up again and starts abusing "cadenas" again.

And in case the reader still doubts, why would a government that genuinely thing that it will win dismiss polls, hire hack pollsters, pack the High Court, set up all sort of electoral tricks, distribute shoddy ID to register questionable voters, suddenly reverse a whole series of policies to try to gain back a sector of the middle class (defense of private property against squatters, new business credits, etc...), and more? Is this is the winner spirit may heaven protect us from the loser spirit!

--------------------------------

(1) I did come across some additional information and edited this post slightly today, July 5, instead of writing an additional one. The information that I found change in nothing my conclusions, but it should be reported so as to provide all the information the readers of this blog need to make up her/his mind.

(2) Since 2001 Chavez has basically stopped giving real press interviews or real press conferences in Venezuela. He only gives them to foreign media and on occasion a tidbit to a particular Venezuelan journalist or a declaration between two doors when he is caught by a journalist. Basically the only messages that the country gets from its president are the "cadenas" and "Alo Presidente" where he rants at will with no one able to question him or even ask for more details.

(3) Thanks to Steven for pointing to me through Haloscan that sanctioning those who signed could have become quite a factor in the results of polling questionnaires. Pollsters might have ways to go around this problem, but I personally think that they have not quite solved this.

(4) Numbers from El Nacional Sunday edition (courtesy of Jose Mora Blog for those who want to read the whole article in Spanish) or from El Universal through the week. Pollsters do not publish numbers in general as these are property of their clients. The comments are mine.

(5) Venezuelanalysis rushes in with a second article to take advantage of these recent polls. Some of the data I publish here is put in nice graphs there. However, interestingly they do not elaborate much on the regional poll of Mercanalisis, which is probably the most solid of the whole lot! Oh well...