Brazilian presidential and general elections 2022 (1st round: October 2nd, 2nd round: October 30th)

[Adam Griffin]

Bloomberg on what markets expect:

An outright victory for Lula would surprise markets, wrote Alberto Ramos, a Latin American economist at Goldman Sachs in a research note. Lula getting less than 40% of the votes would also be an unexpected result, he added. Other surprise scenarios tonight would be Bolsonaro earning more than 40% of the votes or less than 30%, and a margin favoring Lula that exceeds 15 percentage points or falls below 5-6 percentage points.

Wow. Groundbreaking insight.

BRB - buying my Bloomberg terminal!

[Adam Griffin]

I wish Jair Bolsonaro nothing but pain and misery.

He's so full of s[inks]t he's already had two hospitalized bowel obstructions as President: I'm sure a third is right around the corner!

[Adam Griffin]

0.16% reporting

Lula 45.7%
Bolsonaro 42.5%

[Adam Griffin]

0.31% reporting

46.76% Bolsonaro (183,108)
41.84% Lula (163,855)

https://resultados.tse.jus.br/oficial/app/index.html#/eleicao/resultados

[Adam Griffin]

Federal District: Bolsonaro +17
Amazonas: Bolsonaro +17
Sao Paulo: Bolsonaro +20

[Adam Griffin]

The vote counted is mostly from Brasilia lol. Relax people

Id have thought Brasilia would be for Lula.

Nah, it was predicted Distrito Federal (Brasilia) would go strong for Bolsonaro, so this is expected.

That seems weird, why would Brasilia go for him Bolsonaro ?

Plenty of places in Latin America (Brazil, Chile, Bolivia, Peru, etc) still have a nasty habit of urban, wealthy, educated European-originated populations leaning right while the poorer, more indigenous and/or minority groups being more represented in rural areas & embracing left-populism.

It's actually not the worst arrangement of political coalitions when compared to, say, the modern US dynamics. Left-wingers can actually still solidly win with such coalitions.

(30,000-foot explanation from an amateur observer)

[Adam Griffin]

Roughly half of counted votes come from these 5 districts:

Rio Grande do Sul
Amazonas
DF
Parana
Sao Paulo

Not sure how this compares to actual expected vote share.

[Adam Griffin]

Roughly half of counted votes come from these 5 districts:

Rio Grande do Sul
Amazonas
DF
Parana
Rio de Janeiro

Not sure how this compares to actual expected vote share.

Well, only two of those are in the top five states for population, and DF and Amazonas aren't even in the top 10.

As of now (1438k)


DF - 422k
Parana - 100k
Amazonas - 85k
Rio Grande do Sul - 82k
Sao Paulo (accidentally included RJ in the list prior instead) - 35k
TOTAL - 724k (50.34% of national)

 

[Adam Griffin]

https://twitter.com/dwnews/status/1576667031956586499?s=20&t=2r-xSkHfNRSM3XG...

I thought there aren't exit polls this year? Is DW confusing it with the beginning of the actual count?

Based on the quote in the article and how it tracked with vote reporting, I think so. It even says it's the electoral commission's count:

The first numbers published by the electoral commission put Lula at 51% support. This was soon reduced to 47%. And then a third update around 20 minutes after polls closed gave a slight lead to Bolsonaro instead, at 47.6% to Lula's 41%.

[Adam Griffin]

10 biggest cities in Brazil:

São Paulo: Lula 50,88% - Haddad 39,62%
Rio de Janeiro: Lula 44,64% - Haddad 33,65%
Brasilia: Lula 36,15% - Haddad 30,01%
Salvador: Lula 62,41% - Haddad 68,59%
Fortaleza: Lula 53,15% - Haddad 55,61%
Belo Horizonte: Lula 42,12% - Haddad 34,41%
Manaus: Lula 35,91% - Haddad 34,28%
Curitiba: Lula 29,76% - Haddad 23,46%
Recife: Lula 55,42% - Haddad 52,50%
Goiânia: Lula 32,79% - Haddad 25,80%

Lula is running ahead of Haddad's second round results in almost all of these, by an average of around 5 points.

Is that average figure margin shift/swing (i.e. Haddad -20 to Lula -15), or Lula's vote share increase (i.e. Haddad -20 to Lula -10)?

[Adam Griffin]

10 biggest cities in Brazil:

São Paulo: Lula 50,88% - Haddad 39,62%
Rio de Janeiro: Lula 44,64% - Haddad 33,65%
Brasilia: Lula 36,15% - Haddad 30,01%
Salvador: Lula 62,41% - Haddad 68,59%
Fortaleza: Lula 53,15% - Haddad 55,61%
Belo Horizonte: Lula 42,12% - Haddad 34,41%
Manaus: Lula 35,91% - Haddad 34,28%
Curitiba: Lula 29,76% - Haddad 23,46%
Recife: Lula 55,42% - Haddad 52,50%
Goiânia: Lula 32,79% - Haddad 25,80%

Lula is running ahead of Haddad's second round results in almost all of these, by an average of around 5 points.

Is that average figure margin shift/swing (of ~5 points), or Lula's vote share increase (i.e. a ~10-point shift)?

Percentage change. So yeah, if you'd put it in swing terms that'd be around ~9 points. But there's no reason to do so since what we're looking at here is percentage, not margin.

I was just thinking that if it was a 10-point margin swing relative to 2nd round prior & Bolsonaro won by 10-11, it's a decent sign that Lula can still clear 50.

[Adam Griffin]

25.83% reporting
47.74% Bolsonaro (+4.25)
43.49% Lula

[Adam Griffin]

33.18% reporting
47.45% - 17,799,016 - Bolsonaro (+3.70)
43.75% - 16,410,488 - Lula

Definitely speeding up: an estimated 7% of the precincts/vote have come in over the past 9 minutes.

[Adam Griffin]

Since my last update (33.18% reporting), 3.48% of the then-remaining precincts have came in - while Bolsonaro's winning margin declined by 5.95% of its previous total.

If even this trend continues - ignoring more favorable Lula areas remaining outstanding - Lula will be in the lead by 2-3 points minimum in the end. Bolsonaro may overperform polling expectations, but it's not difficult to see Lula getting 47-48%+ in the end.

[Adam Griffin]

52.21% reporting
46.31% - 27737249 - Bolsonaro (+1.45)
44.86% - 26873309 - Lula

Roughly 35% of Bolsonaro's margin closed by Lula out of 10% of the outstanding vote (measured from 46.x% to 52.2% reporting).

[Adam Griffin]

One thing I wonder about: currently, precincts reporting (%) relative to counted vote suggests 120.1m voter turnout - but it was 117m in the first round last time and 114m in the second round. Surely Brazil has grown enough since and general polarization effects would lift this number higher? If we're looking at 125-130m turnout, then it's possible Lula has a bit more breathing room to avoid a runoff (though still unlikely, and obviously depends on where any potential turnout surge originates).

[Adam Griffin]

60.33% reporting
46.01% Bolsonaro - 32049235 (+0.81)
45.20% Lula - 31484583

[Adam Griffin]

Looks like Lula should overtake Bolsonaro when 65-67% of precincts are reporting - and if current trends hold, it's on track for a Lula win of 2-3 points (46.5-47.5%). Still reason to think what's left should be more favorable to Lula than what has been counted, so 48%+ is possible.

[Adam Griffin]

On track for 48.1% minimum (assuming what remains is no more pro-Lula than what has been counted, which I assume it will be even more favorable to Lula).

[Adam Griffin]

CNN Brasil, in classic US fashion and dragging its heels, just projected a second round after the literal government calls it, lol.

[Adam Griffin]

CNN Brasil, in classic US fashion and dragging its heels, just projected a second round, lol.

I guess they are still haunted by the USA 2000 fiasco

Literally the election commission called it - that's the only reason why they even bothered calling it this late.

[Adam Griffin]

Looking like a 5-point win for Lula at this point (48.3-43.3). Assuming universal turnout parity in the second round (I know, but bear with it), even just a bit over half of Gomes voters going for him basically puts Lula at 50. That's ignoring anyone else who I'm sure will pad it a bit more.

[Adam Griffin]

Bolsonaro's raw vote is higher than last time. Is this due to higher turnout (as a percentage) or is it because of population growth?

117.4m in the first round in 2018; we're on track for 123.8m (another ~900k to count) this time. Bolsonaro will probably do around 2m raw votes better out of 6m+ extra cast this time.

[Adam Griffin]

Whoever had that 48% prediction did a great job.

With 99,55% counted the Lula lead finally reaches 5%. Whoever said between 5% - 5,5% in the final result was spot on.

While I know jack-s[inks]t about Brazil, it's pretty easy to keep a relatively accurate running projection once a handful of votes start being counted (i.e. serving in place of "the needle" or other ET prognosticators on Election Night anywhere). Perhaps my refined projections will be off margin-wise by a few tenths of a point at most, but whatever: I stand by my efforts. It's not hard to more or less project even early on when you're monitoring every update and have at least a cursory understanding of what's outstanding.

If even this trend continues - ignoring more favorable Lula areas remaining outstanding - Lula will be in the lead by 2-3 points minimum in the end. Bolsonaro may overperform polling expectations, but it's not difficult to see Lula getting 47-48%+ in the end.

One thing I wonder about: currently, precincts reporting (%) relative to counted vote suggests 120.1m voter turnout - but it was 117m in the first round last time and 114m in the second round. Surely Brazil has grown enough since and general polarization effects would lift this number higher? If we're looking at 125-130m turnout, then it's possible Lula has a bit more breathing room to avoid a runoff (though still unlikely, and obviously depends on where any potential turnout surge originates).

Looks like Lula should overtake Bolsonaro when 65-67% of precincts are reporting - and if current trends hold, it's on track for a Lula win of 2-3 points (46.5-47.5%). Still reason to think what's left should be more favorable to Lula than what has been counted, so 48%+ is possible.

66.66% Reporting

BOLSONARO- 35,360,274 (45.69%)
LULA- 35,246,748 (45.55%)

On track for 48.1% minimum (assuming what remains is no more pro-Lula than what has been counted, which I assume it will be even more favorable to Lula).

Looking like a 5-point win for Lula at this point (48.3-43.3).

117.4m in the first round in 2018; we're on track for 123.8m (another ~900k to count) this time. Bolsonaro will probably do around 2m raw votes better out of 6m+ extra cast this time.

[Adam Griffin]

Here's a question for our Brazilians.

The sub judice category of votes (Anulados Sub Judice) which currently shows 0: is this a future matter of addition or deduction? In other words, are votes currently counted for candidates likely to be annulled, or are all votes in this category currently not being reported at all, and therefore likely to surface in the coming days?

If I projected USA standards of how "provisional" ballots work and assuming one of the two above options apply, this could further shift the margin by 100k votes in either direction (to be fair, the current margin with 99.99% of electoral districts reporting is 6185k votes, so it may not be a lot in the grand scheme of things). Regardless of US bias and coalition differences in Brazil, I imagine an addition scenario would help Lula (given I can't see how it wouldn't disproportionately impact poorer and more rural voters), while a deduction scenario would help Bolsonaro.