Brazil Election - 5 October 2014

[Paleobrazilian]

Veja was echoed by both Folha and Estadão today. I think Globo was afraid of showing the material until other news sources confirmed the veracity of the info.

[jaichind]

Ibope

Dilma 49%, Aécio 43%
valid votes: Dilma 53%, Aécio 47%


Datafolha

Dilma 47%, Aécio 43%
valid votes: Dilma 52%, Aécio 48%

This represents a small shift toward Aecio from a few days ago.

[Paleobrazilian]

Ibope

Dilma 49%, Aécio 43%
valid votes: Dilma 53%, Aécio 47%


Datafolha

Dilma 47%, Aécio 43%
valid votes: Dilma 52%, Aécio 48%

This represents a small shift toward Aecio from a few days ago.

A 1% swing. The problem when the sample was done. Ibope started interviewing voters on Monday, at a moment when Dilma reached 54% of the votes. It wouldn't be surprising if it's close to a tie right now due to the most recent corruption accusations and yesterday's debate. It's like a blurred photograph of a moving car.

[arones358]

Ibope

Dilma 49%, Aécio 43%
valid votes: Dilma 53%, Aécio 47%


Datafolha

Dilma 47%, Aécio 43%
valid votes: Dilma 52%, Aécio 48%

This represents a small shift toward Aecio from a few days ago.

A 1% swing. The problem when the sample was done. Ibope started interviewing voters on Monday, at a moment when Dilma reached 54% of the votes. It wouldn't be surprising if it's close to a tie right now due to the most recent corruption accusations and yesterday's debate. It's like a blurred photograph of a moving car.

That's not even the biggest problem with Ibope. Does anyone really believe that Aecio only has 47% in the South (compared to 44% for Dilma)? In their previous poll Dilma was actually winning in the South even though their state poll for Rio Grande Do Sul showed Aecio leading. So even though Aecio leads Dilma in Rio Grande do Sul he loses to her in Parana and Santa Catarina? It makes no sense.

I for one also think that with state polls in Goias, DF and MS its clear Aecio will win by at least 20 points in the Midwest and erase the gains Dilma has in the North. The question is how will Dilma will do in the North East and will PSDB victories in SP, Parana and MG be enough to make up for those loses. I think RJ is another unknown.

[Paleobrazilian]

I've heard many rumblings about both institutes using skewed samples. Ibope is not very transparent about their numbers, Datafolha is much better on this aspect. This guy explains possible mistakes:

https://twitter.com/relances/with_replies

Also, no matter what sample is used, it's usually much harder to poll the 5-10% richer part of the population, so they're usually underpolled. Among them, the PSDB wins by a landslide.

According to Lauro Jardim, PSDB's statisticians and pollsters are adamant they'll win tomorrow.

[Paleobrazilian]

There are many state races going on right now. Ibope released their numbers for Pará, and it's all tied up at 50-50 between Helder Barbalho (PMDB) and Simão Jatene (PSDB). Jatene gained 2% since the last poll so momentum is on his side.

[buritobr]

Ibope can overstimate the percentage of the PT candidates because of incompetence. But not because of the political bias. The clients of Ibope are the conservative media groups Globo and Estado.

[Paleobrazilian]

Actually Ibope has huge contracts with the federal government so they do have interests here.

[buritobr]

http://g1.globo.com/politica/eleicoes/2014/blog/eleicao-em-numeros/post/datafolha-para-presidente-por-classe-social-idade-escolaridade-renda-sexo-porte-do-municipio-e-regiao.html

http://g1.globo.com/politica/eleicoes/2014/blog/eleicao-em-numeros/post/veja-desempenho-de-dilma-e-aecio-em-sp-mg-rj-rs-ce-e-df-segundo-o-datafolha.html

Aécio's rise in the last three days was not widespread equally in the whole country. It was concentrated in the states of São Paulo and Minas Gerais. Maybe, a strong campaign in both states played a more important role than the bad performance of Dilma in the debate.

[Paleobrazilian]

I take those crosstabs with a grain of salt as well because cities in those states/regions are usually randomly chosen, so it's hard to poll whether some areas of the state are being overpolled or underpolled.

Sensus calls it for Aecio, 52% x 48%. Vox Populi obviously gives the most pro-Dilma result, 54% x 46%.

[Paleobrazilian]

In the middle of the electoral drama, the whistleblower that reportedly implicated Lula and Dilma in the Petrobras scandal, Alberto Youssef, has been rushed this afternoon to the hospital from the prison he was being kept after feeling unwell and is now under intensive care. As you'd imagine, it didn't take long until the first rumors about why he fell ill started floating around.

[buritobr]

Details of the Datafolha poll October 25th

Total: Dilma 52%, Aécio 48%

Gender
Male: Dilma 51%, Aécio 49%
Female: Dilma 54%, Aécio 46%

Age
16-24: Dilma 50%, Aécio 50%
25-34: Dilma 52%, Aécio 48%
35-44: Dilma 55%, Aécio 45%
45-59: Dilma 53%, Aécio 47%
60-: Dilma 50%, Aécio 50%

School level
Elementary: Dilma 61%, Aécio 39%
High School: Dilma 51%, Aécio 49%
College: Dilma 39%, Aécio 61%

Household Income
Less than 2 MW: Dilma 63%, Aécio 37%
Between 2MW and 5MW: Dilma 50%, Aécio 50%
Between 5MW and 10MW: Dilma 40%, Aécio 60%
More than 10MW: Dilma 35%, Aécio 65%

Region
Southeast: Dilma 44%, Aécio 56%
South: Dilma 45%, Aécio 55%
Northeast: Dilma 70%, Aécio 30%
Center-West: Dilma 44%, Aécio 56%
North: Dilma 58%, Aécio 42%

Type of municipality
Metropolitan Region: Dilma 52%, Aécio 48%
Countryside: Dilma 52%, Aécio 48%

Size of municipality
Less than 50k: Dilma 57%, Aécio 43%
Between 50k and 200k: Dilma 51%, Aécio 49%
Between 200k and 500k: Dilma 47%, Aécio 53%
More than 500k: Dilma 50%, Aécio 50%

[buritobr]

Anti-PT activists tried to spread hoaxes which say that Youssef was poisoned.

[jaichind]

Prediction

Dilma 51%, Aécio 49%

[Paleobrazilian]

According to early reports Aecio is running gigantic landslides in foreign polling stations. While this was somewhat expected, if turnout among voters abroad is high (over 300000 voters are eligible to vote outside Brazil), and the election is über close, perhaps this is where the election will be decided

[Donnie]

PREDICTION
Aécio Neves    50,2%
Dilma              49,8%

[Tender Branson]

Might as well try a prediction too for this one:

51.8% Dilma (+/- 1% MoE)
48.2% Neves (+/- 1% MoE)

(% of valid votes)

Turnout: 79% (+/- 2% MoE)

[Niemeyerite]

My prediction:

Dilma 51,5%
Aécio 48,5%

[Paleobrazilian]

Polls at the areas under Brasilia time (the Federal District + Goiás + South and Southeast states) close at the top of the hour.

[Paleobrazilian]

Both sides are concerned about abstention.

[jaichind]

Polls at the areas under Brasilia time (the Federal District + Goiás + South and Southeast states) close at the top of the hour.

So polls close 5pm local time right ?  Is the reason why the election is extended out to 8pm Brasilia time because of places like Rio Branco  ?

[jaichind]

Both sides are concerned about abstention.

What is the penalty of not voting?

[MASHED POTATOES. VOTE!]

I'm not confident enough to predict exact percentage, but I'm confident we'll see a very narrow (pyrrhic, one may even say) Dilma's victory within less than 3%.

Which, of course, would make 2014 the closest Brazilian presidential election ever.

[buritobr]

Both sides are concerned about abstention.

What is the penalty of not voting?

If someone did not vote, she has one month to "justificar". That means: go to the office of the Electoral Justice, fill and sign a form.
If she hasn't given this form during the deadline, she has to pay a fine. The value of this fine is aproximately two dollars.
If she hasn't paid this fine, she can not take the passport, work in the public sector and study at a public university.

[🦀🎂🦀🎂]

Prediction:

52.4% Dilma
47.6% Aécio

This has been a rollercoaster