Thursday, August 16, 2012

Venezuela's electoral numbers 2012: 2- the three key states

The contours of the October vote are getting clearer as we advance in the campaign.  Still, as it is always the case in any election, there is what one calls "battleground" or "key states" which are the ones to watch.  I think that this time around the ones to watch are Anzoategui, Bolivar and Lara.  Let's start with Anzoategui and Bolivar, in the Eastern part of the country, where I think that not only the election will be decided but also the margin of victory (1). I am not considering Miranda or Zulia which are a given for Capriles.  If he loses there, there is no point about worrying how he does elsewhere.  Also Aragua should remain Chavez while Carabobo will go opposition, each one with comparable margins thus cancelling each other.  Hence my choices below.

All the graphs are designed from the same CNE, electoral board, data: presidential result of 2006, governor result of 2008 and legislative result of 2010, as the most motivating elections so as to have the lowest abstention number (which as you will see varies widely).  When the opposition goes divided I add it all up (Lara and Bolivar 2008, Lara 2010).

Anzoategui seems to be a state that chavismo lost, and thus unlikely to recover in October, and certainly not with the dismal campaign of Chavez so far.  The interesting thing there is that if the drop in chavismo votes in 2008 could be correlated with the large increase in abstention, that excuse does not hold for 2010. That year the abstention decreased somewhat but chavismo kept losing votes while the opposition gained enough of those to beat chavismo.  That is, there is a real proof that there is a soft core chavismo vote that can be gained by the opposition.  On the other hand, in a state as mistreated by the regime as Anzoategui has been, that it retains 280.000 votes is a testament on how hard core is the hard core chavismo vote.  Let's say that I think Chavez will get these 280.000 suckers in October and that any opposition further increase will come from a decrease in abstention.

Our next stop is Bolivar, where chavismo seems losing ground more and more.  But let's not anticipate and just look at the recent historical trend.

A priori we have the same story than in Anzoategui: a huge increase in abstention is translated in a huge drop in the pro Chavez vote.  However, in 2010 the abstention decreased and the recovered votes seem to have been split evenly between the regime and its opposition.  Again, it seems that there is a soft chavista vote which is migrating toward the opposition. The last two years give no reason for chavismo to hope for a better result than what it got in 2010.  In fact, this time around, the opposition should win that state, even if by no more that 20-30.000 votes.  You have to remember that Bolivar is heavy in trade unions and if those right now are against Chavez, a few grants are enough to turn them back to chavismo, at least in part, at least enough to avoid a further decrease in votes.  Still, the basic damage has been done and Bolivar is not anymore a chavista vote reserve.

In Lara the situation is more complex because there we have a bona fide division of chavismo and if indeed a not insignificant part of chavismo is willing to follow break away Henri Falcon to the state house that does not mean they will abandon Chavez altogether in October.  That is, in this historically rather leftist state, there is a chance that Chavez may retain the 2010 vote and even improve on it.  This being said, the division there has irremediably compromised the status of Lara as a vote reservoir for chavismo: this one will be lucky if it ekes out a razor thin victory.  Still, my bet is that Capriles will carry the state by at least 50.000, making this the biggest possible upset/loss of chavismo which should have never been threatened there.  But political errors have a price and chavismo committed too many in Lara.

We can summarize the impact of these three states in the next slide.

In this slide I simply added the differential between chavismo and the non-chavez vote.  These three states by themselves gave Chavez in 2006 an appreciable surplus of more than half a million votes, around 20% of Chavez 2006 margin.  But if we were to extrapolate straight the 2010 result to next October, through 2006, an idiotic exercise but bear with me, the result would not be 7.3 to 4.3 but 6.6 to 4.9 (a shift of 700.000 votes). Add Miranda, Zulia, Caracas and Tachira to get a Capriles victory.  Nueva Esparta, Merida and Falcon give you the icing of the cake.  Monagas and Sucre make it a landslide.

But I digress.  The point of these graphs is to bring home the notion that to win in October either Chavez or Capriles need to carry AT LEAST two of these three states.  Why? Because they have a voting weight and because they are the most telling of electoral trends since they represent three different political situations that may converge nevertheless in Capriles favor. If Capriles leads/carries the three of them, even with narrow margins, then it will mean that he wins/won the election

1) When I covered the primaries of the opposition I wrote that the result would be played in the Eastern,  Oriente, of Venezuela.  The result did not illustrate it as well as I wanted because the alliance between Leopolodo Lopez and Capriles threw the election for Capriles and hid Lopez progress there.  Still, Perez did not do well in Oriente, indirectly confirming that the margin of victory of Capriles was decided there (never mind that Voluntad Popular of LL did good in the area in other primaries).


  1. Thanks, Daniel. You are the only electoral expert I respect.

    1. thanks!

      and you may provoked the ire of the "anonymous" below :)

  2. Daniel I don't understand it all, but I appreciate the tremendous amount of diagnostic work and thought you put into this.

    1. may i help you further? do you have specific questions? i try to write as briefly as possible but that post could have been easily twice as long and there are details that i should have added.

    2. I get the gist of what you are saying. I don't know as an outsider who CNE is, and didn't quite understand the graphs and where you acquired the data for the graphs.

    3. OK, I will be more careful.

      You need to understand that after soon 10 years, this is a specialty blog which receives the same set of visitors already knowledgeable on Venezuela. Once upon a time I was excruciatingly boring with details because I wanted people to know what a piece of shit Chavez was. That work is done, successfully.

  3. Anonymous11:13 AM

    This doesn't say anything as you are mixing Chavez personal results with the vote for those backing Chavez. Chavez himself has always scored better than the parties that back him.

    1. Anonymous11:47 AM

      "4) This is an anti Chavez blog, with mostly anti Chavez readers that have made up their minds over twelve years. Trying to convince us of our errors will likely be considered as a mere troll."

      I see I am wasting my time with reason. Bye :)

    2. and if you read carefully my posts, and not only this one as i have written many electoral analysis, you would know that i always consider that factor, that chavez scores better than his henchmen.

      but of course, if your only objective is to let us know that chavez is a demi god, then whatever i write will come across as misguided.

      bye :P

  4. Island Canuck12:27 PM

    "But if we were to extrapolate straight the 2010 result to next October, through 2006, an idiotic exercise but bear with me, the result would not be 7.3 to 4.3 but 6.6 to 4.9 (a shift of 700.000 votes). Add Miranda, Zulia, Caracas and Tachira to get a Capriles victory. Nueva Esparta, Merida and Falcon give you the icing of the cake. Monagas and Sucre make it a landslide."

    My feelings exactly. When you see the turnouts in the last couple of days then this scenario is very plausible.
    The biggest problem Chavismo will have is apathy and low voter turnout, especially if the voter lines are filled with HCR's supporters.

    Hay un camino!

  5. Anonymous2:52 PM

    what margin does HCR need to avoid fraud or such a close election that would give Chavez an excuse to do something stupid?

    1. i will write on that soon, as i keep plodding through my numbers. but out of hand, capriles needs 500,000 to make it difficult for chavez to steal the election.

    2. Anon. asks a very good question. In this case, 51% of the actual vote will equal a loss. I think that we will have to achieve a minimum of 58% of the vote in order to make it stick.

  6. kernel_panic6:01 PM

    Daniel, in the case of reddy-red states (such as, supposedly, barinas, apure, portuguesa and delta amacuro), it isnt likely that the oppo is going to win, but the question is: do you think that the oppo can "rasguñarle la arepa" al gobierno (snag votes from chavismo) there in an important amount? I mean, if we lose there it's ok, but it's not the same to lose 3:1 than 3:2 :)

    Now, remembering the electoral maps once I saw thanks to a link on this website (or CC, cant remember :$), states could be classified as
    a) Sure Oppo (Oppo more than 20% advantage)
    b) Sure Chavez (Chavez more than 20% advantage)
    c) Oppo (by more than 5% but less than 20%)
    d) Chavez (by more than 5% but less than 20%)
    e) Undecided (equal or less than 5% advantage for either side)

    For current elections and the categories described above:
    1) Which states would fall under which category?
    2) What is the electoral weight (approx vote tally for each side) on each category?
    3) What has been the dynamic behavior of the categories described above, state-wise and voter-wise (i.e. how many states on each category have there been each election and how many votes does that translate into)
    4) What would be the effect of a 10% increase in oppo vote in category c and 5% in category d? (what my original question refered to :p)

    1. kernel_panic6:04 PM

      Sorry, question 4 has a typo, the question is

      4) What would be the effect of a 10% increase in oppo vote in category b and 5% in category d?

      and 5: what's your take on category e?

    2. This is all complicated and demand a lot of work. I am not an "encuestadora" and I work alone from my computer so I can only percolate some of it as days go by.

      What you are talking about are "reservoir" states, those where comfortable victories allow for defeats elsewhere.

      In this post what I describe is three states that used to be Chavez vote reservoirs but that are not anymore. To qualify as a "reservoir" state you need to not only favor a candidate by more than 55%, but also have a significant vote count. As such half a dozen states where Chavez wins big cannot qualify (Delta Amacuro, Cojedes, Barinas, Apure, Amazonas, etc) because they do not have enough votes (though put together, they represent quite something).

      We have the big states (Caracas, Miranda, Zulia, Carabobo) and the medium states (Lara, Aragua, Tachira, Anzoategui, Bolivar). Reservoir states can exist only in those states, if they go big on one or the other electoral option. Today the opposition has no reservoir state except perhaps Tachira (though Zulia and Miranda are getting close) and chavismo has lost all it had as reservoir except perhaps Aragua. In other words, no one is certain in these states to win by more than a 10% margin which would make these states a "reserve" of votes.

    3. kernel_panic6:25 PM

      Daniel, take it easy, your answer was what I wanted, thank you!

      I know you're not an encuestadora, you're just a guy who has quirky things that you like and you're good at.

      Keep it up!! Much people enjoy reading you and you are spot on! :D

  7. Dr. Faustus12:12 AM

    I found the above postings fascinating. Good stuff.

    Correct me if I'm wrong, but it appears that the Chavistas have marched into this election campaign ill-prepared. Two days ago USA today had a page 2 article noting how empty the store shelves are in Venezuela. How in God's name could the PSUV have negleged something as basic as that? So, like, what? The purchased food is stuck in freight storage containers at Puerto Cabello? Which soon-to-be-fired Chavista was in charge of that? All of the above statistics are interesting to read. But,...but, if there's no food in the store shelves come election day, Chavez is toast. It's as simple as that.

    1. Stefan4:06 PM

      I love the last paragraph of the article in "US TODAY": "Romero, who has voted for Chávez, is thinking of supporting Capriles. `I am sure the president doesn't know what is happening in the countryside,` he says. `I tried to call him but I couldn't get through. This is a disaster.`"

  8. Dr. Faustus12:47 AM

    Published today, Saturday August 18th by EFE

    "The Venezuelan research firm Varianzas reported, at a month and a half from the Oct. 7 presidential election, a virtual tie between President Hugo Chavez and the united opposition candidate Henrique Capriles.

    According to the pollster's projections, if the election were held this Sunday, the leftist head of state would obtain 49.3 percent of the vote compared with 47.5 percent for Capriles, with 3.5 percent undecided.

    The gap of a scant 2.1 percentage points is less that the firm's last figure in June when it stood at 4.3 points, with Chavez leading Capriles by 50.3 percent to 46 percent.

    The survey, based on 2,000 interviews between Aug. 2-15 with a 2.16 percent margin of error, is far from the advantage most polls have given Chavez in recent months, repeatedly showing the president leading Capriles by 15 to 20 points."

    Oh my, Daniel was right. This is a very, very tight race.

    1. Daniel,

      Since Venezuela does not use an electoral college system, and since the presidential vote is based on a simple majority of the popular vote, I kind of question the significance of analyzing the elections by States. Now, I do understand that the data for comparison comes state-by-state, which makes this analysis method the easiest. However, I think that a more useful breakdown would be:

      The three major urban populations (Caracas, Maracaibo, and Valencia) separately, and then the remainder broken down into 3 or 4 regional zones, in turn, broken down into cities of 50,000 and above and the remainder being smaller cities, towns and campo. That would give you 9 or 11 groups to analyze separately, that have similar social and political dynamics.

      Is it even possible to break down the existing data in this way?

    2. I think that breaking it down by state is not only convenient but significant, even though Venezuela uses a First Past the Post vote.

      How each state leans is useful to know, especially because elections for Governor are coming up in December. Later we will have, finally, elections for local authorities like mayors and city councils (if we still have any left)

      How each state leans makes it easier to analyze and to determine which states need more attention. Thanks to Daniel's work we can see which are the three pivotal, bellwether states needed to win (assuming no major reversals in other ones).

      Chavismo is expected to win in states like Portuguesa, Barinas, Guarico and we are expected to win in Miranda, Zulia and in the Federal District (I hope).

      Different results in any of these are signs of important shifts in the layout of Venezuelan politics. Breaking things down in larger areas might be easier on one hand, but also paint a fuzzier picture.

      But hey, I'm not Daniel so he probably has a better answer than me!

    3. Island Canuck3:10 PM

      "The gap of a scant 2.1 percentage points.."

      Let nobody forget the "fear factor" in all Venezuelan polls. Employees of government companies & ministries as well as pensioners, etc. are never going to admit over a phone that they will vote for Capriles.

      At the least this is 10%, maybe even more. Take that number away from Chavez & add it to Capriles. Then add in the apathy of the Chavez vote after 14 years of red shirts & forced attendance at rallies. Just look at the turnout of a Chavez forced rally compared to the volunteer turnouts at Capriles appearances.

      You can feel the change in the country. Chavez will lose big time in the actual vote. The trick will be to make sure the votes count.

    4. Island,

      Eic Ekvall wrote, "In the 2006 presidential elections Penn & Schoen’s Doug Schoen proved that there was a consistent 14-point bias in favor of the Chávez vote built into the results of standard in-home polls. (Forget about telephone polls — there the bias must be around 25 points.) "

      - firepigette


Comments policy:

1) Comments are moderated after the sixth day of publication. It may take up to a day or two for your note to appear then.

2) Your post will appear if you follow the basic polite rules of discourse. I will be ruthless in erasing, as well as those who replied to any off rule comment.