Tuesday, February 21, 2012

The 2012 primary results: part 2, the turnout

The biggest surprise of Sunday 12 was without a doubt the large turnout of the voters for a primary.  This type of elections are usually an affair between committed and not so committed supporters of a given movement.  Thus the turnout is usually in the single digit numbers and can raise, with a good fight, in the low two digits numbers (less than 15% to put a number).  The reason is very simple: in modern democracies the majority party rarely reaches 60% of the final vote and the abstention rate hovers at 30%.  Thus the theoretical ceiling of a primary vote would be a 40% turnout, a number never seen anywhere.  I am not talking about the US extensive primary system because it is now a tradition, because it varies among states, because it is set to provide a winner A.S.A.P. (though of late it has been failing quite a lot at that).  But when I lived in the US I remember political pundits being ecstatic at a given primary getting a 5% turnout!

For national primaries we have the recent example of France where the Socialists elected François Hollande with  less than 3 million participation.  That is, in France, more than twice the size the population of Venezuela, without any constraints or governmental interference, in percentage the turnout was less than half the Venezuelan turnout.  The French political class was duly impressed by the success of Hollande who has been on top of the polls ever since.  And yet the turnout was single digit..........

Thus, on mere international comparisons the Venezuelan turnout of February 12 is remarkable, a record participation say some, the more so that it took place amid constant attempts at sabotaging by the regime who feared a high turnout and thus did all what it could in "acceptable" scare tactics to keep it below 1.5 million voters.  It got twice that much and sent the regime into a tail spin of invective, fascist maneuvers and what not.  But the opposition in Venezuela should not be too happy either.  If it is true that its chances are greatly improved for October presidential vote, it is also true that the vote of February 12 does not necessarily translate by at least an automatic doubling of its count next October as it seems to be the rule of thumb elsewhere. There are two ways to look at that, the intuitive one and the number crunching.

Is the mood for a doubling of the vote in October?

A priori, yes.  And even more.  But the campaign is barely starting and Chavez has already been pouring so much money into the streets to buy votes that the economic effect is already being perceived.  No need to discuss that effect which will create a fake boom impression to leave the winner of October with a nearly bankrupt country.  But the fact of the matter is that the regime is not going to play fair, it is going to be a very, very nasty campaign where the regime will go as far as closing Globovision if necessary.

In the other hand the exasperation of the people against the regime is palpable, in TV, in radio, in your everyday encounters.  That is why in January 11 I was already sensing the high turnout and on the basis of historical voting patterns I went as far as predicting a 3 million people turnout.  It is interesting, by the way, to read today that some pollsters were seeing a high turnout but were not willing to go on record with it on some lame excuses as "no precedent to compare"...  Certainly not increasing my faith in Venezuela's pollsters that always manage to pretend that they were right all along......

But that "palpable feel" is double edged.  It means in part like people like me are tired of the whole thing , so tired of it, that they do not care anymore about being vocal in public.  We are the same people who voted NO in 1999, who signed any petition handed our way in 2002 and 2003, who loudly claimed with pride that they were in the Tascon list, etc, etc....  And yet went on to lose in 2004 and 2006, badly that last year.

However the circumstances are different today and we can sense that in a close scrutiny of the raw numbers already: in 2004 3.98 million voted in the recall election in favor of ousting Chavez.  This number is rather solid, the one where the cheating happened was on the 5.80 we are told Chavez got.  Thus the 3 million of a few days ago is telling of things to come.

All logic, all primary historical experiences, all 8 years of Chavez tenure since 2004 tell us that among the 3 million that voted Sunday 12 there were quite a few that did vote NO in 2004.  Heck, just start counting already all the chavista political figures of 2004 who now were running in primary elections for local MUD nominations.  Thus even if we consider it all, the 2 X rule of vote multiplying from primary to general election, the 2004 precedent, the "Firmazo" of 2003, etc, etc... it is a fair bet to say that of the 3 million of votes cast two Sundays ago, up to1 million may come from the new, post 2004 opposition.  Thus comparisons between processes so spaced in time should not be made.  But let's do them anyway.

2003 signatures versus 2012 primary
In 2003 we were asked in December to sign up for a recall election on Chavez.  3.6 million of us did, but the CNE eventually after a long protracted battle had to approve the signature process but managed to erase1 million names.  So, with 2.6 million signatures we went on to cast 4 million votes.  You could say that the X 2 rule was sort of respected there.  However think that conditions were different: with the existence of the Tascon list then in full swing.  It is quite certain that many voted NO in fear of being discovered as the misiones were also in full swing there and no one inside the "coordinadora democratica" was proposing maintaining them.  Chavez had gained 6 precious month of outrageous populism that changed dramatically his polling numbers.

I went back to look at these numbers and I made the table on the right comparing the outcomes of Sunday 12 with the initial 3.6 signatures collected as given to the regime then.  Certainly the regime never recognized the 3.6 million signatures but ALL found their way into the Tascon list now called Maisanta once the pro Chavez people were included depending on what mision they belonged to.

For simplification I have written the tables of this post all in thousand (K) of signatures or votes, and rounded it all up for simplicity.  Since we are talking estimates and speculation here, decimals are of little use except in the turnout numbers.  The states are ranked according to their turnout Sunday 12 (see below table).  Miranda with Capriles as its governor was of course the state with the higher turnout, while distant, dependent and inaccessible Delta Amacuro was the one with the lightest turnout.  It is thus not unexpected that those states were dependency on the misiones and other social programs are those who voted the least, and even at half the rate of the signatures of 2003 (bold red).  Those states have probably developed quite a network of informants and thus only people that are independent form state resources can afford to be seen going to vote.  However at least two states that were heavily pro Chavez in 2003 did manage to have significantly more people voting this year than signing up in 2003!  Note: if we retain the CNE figure of 2.6 million real signatures in 2003 then the whole interpretation goes in our favor, but I like to be the devil's advocate.

In other words, driving a direct correlation between the firmazo and the primary as some have done is a cheap gambling, although the 2 X factor observation is valid.  In 2003 people were less afraid and more combative.  In 2012 people are tired and more afraid but also more pissed with Chavez unending tenure.  Maybe not apples and oranges but certainly oranges and grapefruit. The 6 million in October is thus quite a possibility and the idea of it will keep Chavez sleepless in many nights to come.

Turnout, percentile

The numbers game

With the provisions set above, it is time to look at how meaningful the actual numbers are.

The graph on the left is a clear illustration of the opposition momentum for October.  This is sensed, but a good representation of numbers help visualize that beyond the punditry found everywhere.  The states were ranked according to the turnout, from the less to the most.  Note: yet, there is not a fully complete publication of the total results so this table comes from me piecing out numbers from two sources, just in case some one somewhere has the real exact numbers under his or her elbow, unpublished.

In red are those states that are still in chavista hands.  Unsurprisingly they are the ones with the lowest participation, due to 1) fear and 2) that there are more Chavez supporters.  In blue we have the states that went opposition already in 2008 regional elections.  And in red to blue those states whose governors defected to the opposition (Lara and Amazonas) or those who gave the opposition a majority vote in the legislative of 2010.

I think comments are not necessary on that chart: quite clearly the momentum is on the side of the anti Chavez opposition and there is no reason why this should stop before October.  We can almost see the list of the states where Capriles is more likely (blue and red blue) to win to which we can add that he could carry Bolivar and Sucre.  That would be enough for him to win, handsomely even.

However this chart also indicates us the extent of the  X 2 factor for the expected vote in October.  Let's take Miranda with a 25.1% participation.  Can we really assume that Capriles will get 50.2% of the registered voters?  With an expected, say, 25% abstention, that would leave Chavez with LESS than 25% in Miranda!  True, a 2 to 1 victory for Capriles is possible in Miranda but my advice at this early in the campaign is to ease up on our expectations....

Which leads us to the next table.  In that one I decided to increase the turnout according to each state.  That is, I assumed fairly that in Delta Amacuro the turnout in favor of Capriles could increase by a 3 X factor, but in some other states by a much more modest factor.  Then based on the registered voters for this election I calculated the votes that Capriles can gain according to two scenarios.  In the first one I consider a moderate turnout gain.  That is, I assume that Chavez campaign will be successful and that Capriles will "only" get a bounce from voters too afraid of voting a week ago but that would do so in October.  In other words, the lowest voting Capriles should get if his campaign collapsed, something difficult to believe since so far his strategy has been excellent.  With the second scenario I simply push a little bit more the envelope and assume that indeed the "bland" approach of Capriles does succeed, not in prying votes away from near fanatical chavismo, but at least decide the undecided absentee voter to go and vote for him.  Transactional voters, so to speak.

Possible scenarios (note: Venezuelan Excel sheet so dots are commas and vice versa)

First column is the different states classified according to February turnout.  This is read in the third column. The second column is the registered voters in each state.  Keep in mind that these numbers will vary some as the CNE is allowing more people to register, and that there is a strong overseas push to register Venezuelan voter there, making it a "state" comparable to Delta Amacuro in importance, not to be forgotten in the case of a tight election.  Which leads us to the fourth column in blue, the votes cast on February 12.  As you see, my rounding up and lack of absolutely trustworthy information did not prevent me from getting a total of 3 million 50 thousand, which is a few thousand short of the final tally.  I am good so far.

The 4 and 5th columns are my "moderate" scenario.  In it I assume that Delta Amacuro participation in favor of Capriles will go by a 2.5 X, but the one of Miranda only 1.5 X.  here we get a 5.3 million for HCR in October.  It is almost safe to assume  that this is what will happen in the worse case scenario.  And it scares Chavez because it is already 1 MILLION more than what Rosales got in 2006!  Chavez then got 7.3 million and it is difficult to imagine him getting more than that this time around, no matter how much cheating he does.  In fact, he should lose a million votes because this term has been so much worse than his preceding term.  And this is what leads us to the next scenario.

In the last two columns, with a slight shade of hope green, I push the ante some.  Now Delta Amacuro is 3 X and Miranda almost 2 X.  A 43% for HCR in Miranda with a 25% abstention give a more realistic 32% to Chavez there.  I am thus not exaggerating much and thus the new putative total for HCR is 6.3 million, something very reasonable, something quite possible to reach, no matter what Chavez throws at him form vote buying to actual sabotage.

And this is where trouble starts because if Chavez indeed loses 1 million from 2006 then we have 6.3 million for each one of them!!!  For the first time since 1999 Chavez starts an electoral campaign in an even term with Capriles (something already pointed out by Consultores 21, which is a good way to validate my model above, thank you very much).

No wonder Chavez has been in such a fury last week!  If anyone in Venezuela crunches numbers better than most it is him, something that I have recognized long ago in these pages


  1. 1979 Boat People5:10 AM

    If it is true that its chances are greatly improved for October presidential vote, it is also true that the vote of February 12 does not necessarily translate by at least an automatic doubling of its count next October as it seems to be the rule of thumb elsewhere. There are two ways to look at that, the intuitive one and the number crunching.

    You nailed it, Daniel.

  2. Anonymous2:02 PM

    I think another push for the oppo will be the unexpected absences from Chavez. Have not heard or seen him since Saturday. I think this will turn some soft core Chavez followers that are still on the fence.

  3. You know, one thing just occurred to me, which is very positive. With the NA process, there was ridiculous amounts of gerrymandering - both assigning oppo voters to districts where they were overrepresented, plus assigning greater numbers to those districts compared to pro-Chavez districts (and even within the same provinces, never mind the distribution among the provinces).

    None of that matters for President. If it were a U.S. system, with an electoral college, it very well might. No such luck for Hugo (unless he invents it in the next 6 months).

  4. Thank you Daniel, this was the post I was expecting. I am not disappointed!

  5. and I very much agree :)

  6. Anonymous8:08 PM

    We wont win with 6,3millions.

    If the REP is let say 19millions (Right now it's 18,4millions), and we have a turnout of 75% (Like 2006), we have 14,25millons votes.

    So we need at least 7,15millions, that is the number.

    For example Liborio already won the governation with more than 32k votes. Your optimistic scenario is not that optimistic in that state.
    The same happens with Barinas and Tachira (We already got more votes in the 26S 2010 election than your optimistic estimate).

    1. I never said we would win with 6 million. It is a starting point for the campaign.

    2. And if we go by your reasoning then we do not win with 7.15 because it is too close and Chavez can cheat easily. So, let's take it one step at a time, shall we. And this first step is merely to understand the current scenery. The next step in the following weeks is what to do about the numbers offered above.


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.