The Gerrymander By the Numbers
Yup, I get pretty obsessive when it comes to election-related data. At great cost to my sleep rhythms, marital peace and general hygiene, I've spent the last 48 hours trying to make some projections for September's elections to the National Assembly.
Specifically, I was keen to try to get a specific, circuit-by-circuit estimate of the effects that gerrymandering is likely to have on September's election. We all know CNE set out to screw us by redrawing the AN's circuits' boundries, but how well did they succeed?
Here's the gist of it:
.gif)
As expected, the map is rigged. The circuits CNE came up with mean the opposition would need to get north of 52.7% of the popular vote to get a simple majority in the National Assembly. In fact, there's a very real chance that the government will keep control of the Assembly on the basis of a minority of the popular vote. Lindo, ¿no?
Assuming the proportion of chavistas-to-oppositores within a given parroquia is relatively stable over time, I estimate that if the opposition wins a 50%+1 vote majority of September's popular vote, the government would still win a whopping 35 seat majority in the AN: it'd be 100 chavistas to 65 opositores. And the government could get a 3/4ths parliamentary supermajority on as little as 55% of the popular vote.
So the map's certainly rigged, but then the country as a whole is rigged, so that's not really news. To my mind, what's interesting is that CNE wasn't really as aggressive as they might have been. If they'd really put their minds to it - if, say, they'd carved up crazy circuits that cross state lines or split parroquias in two - they could've done much better...by which I mean much, much worse.
As it stands, it's not unimaginable that the opposition could turn the system's quirks to its advantage. As we've discussed before, the socio-economic fundamentals are just plain awful for the government this year: I mean, you know Chávez has it tough when the country's going through stagflation and that's not even one of the three strikes he's ponchao for. So it's not beyond the realm of possibility that the chavista machine could sputter, chavista turnout collapses, the protest vote runs away with it, and the opposition turns 56% of the popular vote into 63% of the seats in the AN. Imagine that?!
Neither can I.
Anyway, there'll be much more Excel-based election tomfoolery in the days ahead...I'm on a roll here...
Oh, and if you want to look over my shoulder as I make these estimations, you can download the Excel sheet I'm using here. It's...um...not for the faint of heart.
[These scenarios are on the basis of the lovely 2007 voting-center-level dataset amieres was good enough to get for me all those years ago...so hat tip amieres!]
Side effects of xanax.
from Buy xanax. on Tue, 02/16/2010 - 02:36Buy xanax. Generic xanax. Weight gain while taking xanax. Xanax.
Free porn video clips.
from Free porn no credit card. on Mon, 02/15/2010 - 04:04Free adult porn. Free porn pics. Free porn pictures. Free porn. Free incest porn.
Prozac.
from How does prozac work. on Sun, 02/14/2010 - 06:24Hair loss prozac. Phentermine prozac. Prozac nation movie. Prozac and pregnancy. Prozac side effect. If prozac makes depression worse. Prozac.
Casino royale 007.
from Casino royale. on Sun, 02/14/2010 - 02:58Soaring eagle casino. Grand casino. Casino royale. Casino. Casino jobs.
Order vicodin.
from Vicodin. on Sun, 02/14/2010 - 01:35Buy vicodin without script. Vicodin and hydroxyzine. Vicodin. Codeine vicodin. Vicodin without prescription. Vicodin addiction.
billige unterkunft buchen
from billige unterkunft buchen on Wed, 02/10/2010 - 20:53[...] Miss America is not supposed to portray the innocent bathing beauty anymore _ she's a career woman who is educated and articulate, and she may not even be a virgin, says the pageant's new director. Life,consumer On Aug. 25, a public council was f...
hotel billig
from hotel billig on Tue, 02/09/2010 - 10:55[...] Trading was moderate, with first section volume estimated at 900 million shares, a large increase from Monday's 550 million. Again,what All the old limits on deductions remain unchanged. Again,what [...]
Social comments and analytics for this post
from uberVU - social comments on Tue, 02/02/2010 - 20:13This post was mentioned on Twitter by CaracasChron: En septiembre, con el 49% de los votos, el chavismo se puede quedar con el 58% de la AN!! http://bit.ly/d9CBDf
.jpg)
.gif)
.jpg)
Lovely
The opposition will get ... 52.6% of the vote, and 35% of the seats. Tibisay has her work cut out for her.
Nice work Quico.
So much gerrymandering and so little silence from the opposition
I suppose they don't want to scare people away from voting. Now they really need a lot of good thinking and listening to all possible groups that come with positive criticism.
Good work, Quico.
More realistic goal, perhaps?
Wouldnt it be TOO ambitious to hope that the oppo gets 53% of the votes?
I mean, if the oppo gets 48% of the votes for 36% of the seats in the AN, wouldnt that be enough to at least throw a wrench in the gears of the chavismo blank-check machine that has become the AN?
From 2003 to 2005, Chavismo had a majority of ONE
That's right: following the string of dessertions caused by the 2002-2003 crisis, the government had 83 votes to its opponents' 82 (or sporadically it was 84-81, depending on which side of the bed Rafael Simón Jiménez got out of that morning.)
And what did they do? They changed the A.N.'s rules of order SEVEN TIMES to strip the parliamentary minority of any ability whatsoever to even slow down the legislative juggernaut until, by the end, the 82 opposition deputies were basically window dressing.
On the broader point, I don't think 53% is such a crazy goal. At the end of 2007, with oil prices in the triple digits, the economy in the middle of a feverish consumption boom, and most people's real incomes rising, we got 50%+. Oh yeah, and the lights were on back then. And the water.
So thinking we can do 3 points better this year than we did in those circumstances doesn't seem all that far-fetched. I mean, coño, we're in a stagflationary spiral. If there's no protest vote now we might as well call it a day.
two points...
Interesting work, Quico, thanks.
But I think it's necessary to consider two different phenomena:
1- Over-representation of rural circumscriptions.
In our old bicameral system, we had one assembly (congreso) with more and less proportional representation of population's vote, but a second one (senado) where rural states were over-represented. If you want to keep this political balance in an unicameral parliament, you must, by construction, give over-representation to rural zones. Constituyentes, in 1999, did it.
The problem: "Opposition" decided to retire from rural zones, and Chavismo have (very) easy strongholds.
2- Mixed majoritarian + proportional electoral system.
Differences between popular vote and seat is absolutely common in all majoritarian electoral systems: for instance, Labour took 55% of seats with only 35% of votes in 2005 in UK (http://en.wikipedia.org/wiki/Majoritarian)
Our system is mixed, so, what is the real effect of consolidation of strong oppositions majorities in small sectors (Chacao-Baruta-El Hatillo, Valencia-north, etc.) with progressive erosion of Chavismo in large popular sectors?
For instance, if opposition score in Chacao changes from 72% to 84%, there is not incidence in the number of seats, as well as if chavismo score in 23 de Enero fall from 61% to 53%.
In any case, CNE's gerrymandering seems to be cosmetic. If you want assembly majority, you need strong presence across the country, including countryside and working-class areas.
Mucho trabajo pendiente!!
There's some truth to that, but PR is a constitutional mandate..
It's easy to lose sight of it, but the 1999 constitution - sabes, "la mejor del mundo" - is explicit in calling for proportional representation. That was the whole point of balancing the circuits with PR lists.
I should note that, even now, the list system works to the opposition's advantage in rebalancing the rural vote to some extent. Rural states each have 2 "list" Assembly members. Under the D'Hondt PR method, the opposition only needs to reach 33.3% of the vote in a given rural state to get at least one of them.
The problem only comes if the bottom REALLY falls off of our rural vote - as it did in Apure, Portuguesa, the Delta, Cojedes and Amazonas in the 2009 referendum. We didn't even clear 30% in those places in 2009. Which, if that's repeated this year, would hand BOTH lista deputies to the government in those states.
But 33.3% is a reasonably achievable goal for the opposition. In 2007, we cleared that threshold in every single rural state.
Actually, one projection I've been twiddling with is something like "the opposition's path of least resistance to 83 seats". Watch this space in the days ahead.
Witnesses and the actas
Delta Amacuro:
The oppo scored in the same parroquia almost 10% better in schools where we had actas. In parish Jose Vidal Marcano:
No vote
school without an acta:
33.97, 25.53, 37.26, 16.17, 3.16!!!, 44.33, 36, 53.38
(casa comunal, perhaps there was a priest or something)
school with at least one acta:
58.24, 50.23, 51.12, 33.72
It is not absolute, but a clear pattern.
Example: 2009, NO vote
Pedernales municipality, Pedernales parish:
escuela Samuel Darío Maldonado:
mesas WITHOUT actas in hands of oppo:
mesa 1: 9.48%
mesa 5: 6.8%
mesas WITH actas in hands of oppo
mesa 2: 13.48%
mesa 3: 20.3%
mesa 4: 15.92%
Even in Carabobo:
In Libertador municipality (250000+ inhabitants) as well as huge Miguel Pena parroquia (500000 inhabitants), where the greatest gerrymandering took place in that region, we urgently need lots of witnesses. You have to see the data from Sumate and Esdata.
We actually have enough people in Northern Valencia but almost everybody in Valencia wants to be a witness in their nice neighbourhoods. There are fewer people wanting to be witness for us in the South not just because of lack of interest: many are afraid.
So we ended up having people to be witness, to be suplemente 1, 2, to pass the cachitos, to keep the cachitos warm at every mesa in Prebo (a part of Northern Valencia)
and very few in the South (the poor).
This we cannot afford anymore.
Another question
This might be a little bit off-topic, but I was hearing from my uncle that, if the oppo wins enough seats in the AN to bother chavez, he could (legally!) dissolve the AN by any excuse (he needs one, but hey, what havnt we seen here? plus, AFAIK, Fujimori did something like it) and call on elections again, is that true? if so, I dont know... what do you guys think? :P
Dissolution of the Assembly
Legally speaking, your uncle is right. According to article 240 of the Constitution, the President may dissolve the Assembly if the Vice President has been removed three times by the Assembly. But this requires three fifths of the votes, which is highly improbable. And what Assembly would want to commit suicide over such an unimportant office? However, there are other "tricks" that Chavez could use: call for a new Constituent Assembly with new rules and districts to ensure a majority for the government; promoting an institutional crisis to provoke gridlock and a solution by the Supreme Court, etc.
habilitate THIS, pal!
My scenario: the outgoing AN approves a massive, sprawling Enabling Law good for 60 months just before the oppo majority is seated. Such an elegant solution!
they could srike it down
Any Habilitating Law could be struck down, but then again, that could be reversed by the TSJ.
It's weird how nobody in the opposition discusses these doomsday scenarios.
Not scaring voters
I think that as someone mentioned somewhere in this blog, they don't want to scare voters, remember that the previous solution for the fraud was abstention and we all know where that took us. Already a lot of people feel is useless to keep voting and if you denounce all the tricks the Chavistas are doing then more people would decide is not worth to go out and vote and right now we need to get every vote… I guess that's why the opposition is focusing on the selection of candidates and not on the CNE marramucias...
Constituent Assembly
I was too young when that happened :p
In fact, I was in 5th grade when Chavez became president, Im about to finish univerity and Chavez is STILL the president... its like pokémon, Ash has been 13 year old for the last 10 years... which makes me think what kind of pokémon chavez is :p (or if he's trying to CATCH US ALL!)
Anyways, regarding the Constituent Assembly, what would need Chavez (or chavismo) to do in order to call for that?
By decree
Simply sign a decree and that's that.
That's with 2007 escenario?
Where are you taking numbers such as our winning in Catia or our winning in Guarenas-Guatire? I think it is highly probably the opposition may scratch one of the 2 seats there, but I dont think we may squeak out... Besides, I think you are really optimistic in Aragua, Yaracuy and Sucre. I hope you are right though.
2D Results
The central scenario is made simply by asking what would happen in each of the just-announced circuitos if people vote exactly as they did on Block A of the 2007 Constitutional Reform Vote. My opinions don't actually figure in the analysis at all.
the spread
quico
i am certainly glad to read that after hours of work accumulating body odors you still reach the same conclusion as i did based on my circumscriptions political knowledge: the opposition needs to take 4 points above chavismo to have a 1 seat advantage.
in other words, whether we look ar it from math or politics, the CNE did a hack job, and surprisingly not as much as it could have done. then again chavismo is used to work with a one digit majority if needed, and thus that is their sole goal.
slanting elections in more ways than one
Daniel,
Agreed.I remember reading about this and also my own comment, some months ago.
"the CNE did a hack job, and surprisingly not as much as it could have done."
This is because Chavismo has more tricks up their sleeve to ensure they " win " the elections.
They don't want to make it too obvious by relying on just one technique.
Assumption
Quico, I have a clarifying question. How do you reconcile your assumption of constant chavistas-oppositores ratio within a given parish and a varying opposition share of the popular vote? I don't see how.
I just mean that the swing is proportional
...so if Circuit A was 48% oppo in 2007 while Circuit B was 55% oppo in 2007, and we want to project a nationwide swin of 10% toward the opposition in Sept. 2010, then I assume the oppo vote rises 10% in EACH circuito...
So circuit A is projected to come in at 52.8% oppo (48 + 4.8) and Circuit B at 60.5% oppo (55 + 5.5)...
A lesson
I've been playing around with the first spreadsheet you posted. One conclusion I've reached is that the opposition doesn't need to spread evenly its efforts across parishes to win more seats than chavistas in the National Assembly in any given state. Basically the opposition has to focus its resources and campaigned hard in the more populated parishes where it has lost (especially if by a small margin). For instance, it makes no sense to spend money in campaigning in places like Chacao or Valencia-San José. The return to that "investment" would be really low, not to say zero. Instead, the expected return to campaigning hard in places like Vlaencia-Miguel Peña and Libertador (Carabobo) is huge because winning Carabobo's circuit 5, where these parishes belong, would render 3 more seats to the opposition and, obvously, 3 less to chavistas. Of course, if the opposition think that the chavista advantage in any particular electoral circuit is unsurmountable, then they should keep their campaigned there at a minimum (I guess the opposition should have a minimum presence everywhere).
All this implies that your estimated 52.7% opposition target might be flawed and meaningless. Let's look at Carabobo. Taking 2008 results, the opposition with 51.6% of the votes takes only 3 seats (out of 10). I ask the following question now: By how much does (x%) the oppo vote need to increase in order to get a majority of the seats? (That's assuming it increases x% in all parishes, and that chavista votes go down accordingly so that the total votes don't change.) I found that x=8.2% which means that now the opposition gets 55.8% of the total votes in Carabobo, good for 7 seats. I'm guessing that my 55.9% for Carabobo is equivalent to your 52.7% for the whole country. Am I right? Now let's ask the following question: By how much does the oppo vote have to increase in Carabobo's circuits 2 and 5 to win 7 seats? (That's assuming that in the remaining electoral circuits the results are as in 2008). I found the same x=8.2%. However, that implies that the opposition gets now 53.4% of total votes in Carabobo, which is still good for 7 seats. I suppose you get the picture and why I think your threshold of 52.7% is flawed.
What do we learn from this? Let's assume first that the opposition has Y Bs. available to campaign in Carabobo. I think it's easier to increase the oppo vote by 8.2% in circuits 2 and 5 (which seem to be the key circuits for the opposition) by spending a very high fraction of Y in those two circuits alone than by spending Y more evenly across the five Carabobo's circuits.
Manuel, you're absolutely right!
What you're doing now is what I intended to do over the next few days. Of course it's artificial to assume a linear swing. But remember, los rusos tambien juegan! (i.e., if we can figure out to focus our firepower in the "Battleground Circuitos", so can they.)
Email me. Lets be Excel BFFs!
Sure the Russians know!
Of course they know...but honestly there is only so much the opposition can do about that...its dominant strategy still is campaigning hard in the big parishes where the race is tight. Plus, they also knew that underinvesting in the electric sector would bring problems and they still didn't do anything! Although it seems that in electoral matters they look a bit more capable.
Biggest Bang for the Oppo Buck
Quico, Manuel,
What Manuel mentioned is exactly what I've been working on last two days. I based my results in the 2008 Regional Elections, specifically in Governor vote at the parish level, which is IMHO an electoral event better fitted for the exercise than the 07' referendum (local enough, yet influenced by national trends) Tomorrow I'll send the results.
Re
I am affected because of the hot stuff about this topic. You done this as the freelance writing , therefore, you deserved the highest price for that, I think!