Monday, April 22, 2013

The April 14 2013 result: the Caucagüita story

I was wondering whether to analyze the electoral results of last Sunday. After all, there is a fraud. But then again, I've never taken CNE results truly at face value and only found their worth in indicating trends away from chavismo, stuff that even manipulated data cannot hide. So I have decided to make only two posts, this is the first one, retaking my poster community, Caucagüita.

I  have chosen this district inside Sucre, the big district of Eastern Caracas, around Petare, dominated by what is perhaps the single biggest "favela" of South America.  Caucagüita was made noteworthy when it was a mere squatter settlement in the campaign that elected Luis Herrera Campins in 1978, where he had one of the most remembered political ads ever (with Aleida Josefina).  Things have changed a little bit now, a highway goes through it and a few high rises have been built. Still, this is an area where chavismo should NEVER get less than 3/4 of the votes cast, and this in a very bad day. I know, I already wrote all of this but I am saving you time here. You are welcome.

Well, if you look at the graph below you certainly can understand why chavismo is getting really, really nervous.
Sorry, my Excel is in Spanish. And I do not want to put the date of the election because it overburdens the graph. In purple light the trend line for the deeper purple of opposition percentile.

If you graph all the results since 2004, and include the best ones when there is to chose between governor or mayor (2004 and 2008), you can observe the following:

In blue the total votes of the opposition is ALWAYS growing. And when it falls, so do fall badly the chavismo vote. The blue line is thus quite steadier than the red line.

Chavismo always wins, which is to be expected, but its vote is erratic and always drops when Chavez is not directly on the ballot.

But the story is in the purple line which is the percentile the opposition gets. Its low is on October 2004, in the post recall election funk. Then it was around 25%, the maximum it should get in Caucagüita. The surpise is that since that result, the percentile has tended to grow when one would have expected that Misiones would have secured that district.  Last Sunday, not only Capriles recovered the October 2012 result but Maduro failed to recover the Chavez one. The trend line is now apparently in an unstoppable way to reach 45%, which for chavismo is simply a disaster that a popular sector can abandon it that way.

This graph, look at it carefully, is the best example on how the opposition has an increasing appeal to the lower classes that are getting tired of been taken for granted by the regime, lower classes which are starting to sense the client role chavismo wants to perpetuate on them. And it is not to their liking.


  1. Anonymous5:42 PM

    Where is the slope of the line for the red line?

    1. There is no need for a slope {trend line, I assume you meant} in the red or blue line. As the population in this sector keeps increasing, both trend lines will be positive. Trend line in total votes are meaningful only when a party starts from zero until it reaches a stable percentage so you can evaluate the growth speed; or when you are discussing a stable district, with little population variation.

      In 10 years Caucaguita has gone from about 17.000 votes cast to 25.000 which is a lot in statistical terms for ten years and as such it is more than likely that both side experience an increase of vote, as the graph indeed shows. That is why it is more interesting to look at the seesawing of the chavista vote and the percentage share of the opposition vote, as I represented.

    2. Anonymous5:08 AM

      Sorry, Daniel. I don't buy the argument. Indeed, the red line has more zigzag, but what is the big deal about that? To compare the trends between the votes for the government and the opposition you should compare the slopes of the trend lines (the two lines start at different elevations, so no direct comparison is possible). At first sigh, I see no evidence that the trend for the growth of votes for the opposition is any stronger, despite the largest zigzagging in the pro government vote.

    3. For somebody who signs anonymous you have quite a lot of nerve in your demands.

      What you are suggesting is to compare apples and oranges. There is a difference in the total vote count that each side get and the proportion that each side gets. If you cannot understand something as basic then there is no further need to discuss this issue.

    4. PS: by the way I cannot fail to notice that so far the only thing you have aimed at is to annul my interpretation but I do not see an interpretation of your own. And yet all the data is there, straight from the CNE.

  2. Anonymous9:05 PM

    How much of the voter increase is due to youth turning 18 years old? How are they effecting the trend?

    1. I doubt there was any exit poll in Caucagüita. Also, the growth of Caucagüita is not only youth.

  3. Anonymous6:26 PM

    Sorry Daniel, this is Pelao. I sign as anonymous because I am unable to sign in with my world press account here. I don't really understand why.

    Let me see if I can make my point clear: If I understand the post correctly, you are arguing for an increasing trend in the percentage of votes of the opposition. That's abundantly obvious. My point is that there is also an increase in the percentage of the pro government vote. This is in itself interest because it has to suggest that abstention must always decrease with time (otherwise, if you essentially have a two-way contest, how is it that both contestant increase their percentage of the vote?). Anyway, Since the percentage of votes is almost always higher for the pro government vote, it seems to me that the only way to see who is getting ahead is to compare the slopes (percentage of the vote per year) of the trend lines. If the lines are parallel, or the pro government line is steeper, then your interpretation and conclusion on this post is on shaky ground. At first sight, the lines for both camps appear parallel to me (and the zigzag is not a relevant issue, unless it makes the slope zero). That is, no one is getting ahead.

    You know, you are a scientist, and so am I. I am interested in testing hypotheses. I am not interested in being married to one. If my hypothesis is incorrect, so be it. You have the data, and can show me that my concern is unfounded. I hope my concern is clear now.


    1. Pelao

      Again, you are comparing apples and oranges.

      The red line is the TOTAL votes, NOT the percentage. So is the blue line, it is the TOTAL votes gotten by the opposition, NOT its percentile. The purple line IS the percentage of votes obtained by the opposition. If I were doing a percentile yellow line for the chavista share this one would go down. Slowly for sure, but surely the yellow trend line would move toward 55% in a year or two from now.

      The percentile line is not quite exact because there are other candidates but their total never amounts much more than a couple of percentile points so I have ignored them because they do not affect the interpretation. That is, to calculate the percentile line of the opposition share I do: (Total oppo vote)/{(total oppo vote)+(total chavista vote)} which means obviously that if my trend line for the oppo goes upo the trend line for chavismo goes down. I did not put it up because the graph was already loaded enough.

      As for the interest of the zigzag. A trend line can be made from any zigzag, the only thing that is affected is its R2 value and its relevance. Another reason not to put a trend line in both oppo or chavista TOTAL votes because both go up!!!! Your intuition is right on that. I did calculate it in a graph not shown and they were almost parallel indeed. So I did not put them. What is interesting is that chavismo is the one with the zigzag: when Chavez is not on the ballot chavismo TOTAL vote always goes down. The opposition does not show such zigzag because new voters come to it from conviction, from realization that the chavista model does not work and thus they are more likely to be more constant voters.

      I hope this is clear now. I can send you the date if you want to play with it yourself.

      As for signing: just type pelao at the end of any comment, you do not need to register in Google or whatever. Yes, I know, it is a pain in the neck for readers to type the little anti robot window, and wait for moderation after 24 hours, but it avoids a lot of junk mail and also fly by chavista comment bombers abstain, preferring to go and harass other blogs that shall remain nameless.

  4. Anonymous2:00 AM

    OK, thanks, Daniel. I had not realized that the red line was the absolute vote, not the percent.


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.