Brazilian Election Final Stretch

In an election filled with twists and turns, Social Media Influence (SMI) seems to have one last final surprise left in store.  A late surge by Ciro Gomes in SMI, which has not been detected by any traditional metric, will likely result in Gomes significantly outperforming polls tomorrow.  Though unlikely he can make up the double-digit percentage point poll deficit against Haddad, SMI points to Gomes considerably closing the gap.  In other words, second place is not as secure as most believe.

According to SMI, Ciro Gomes won the final televised presidential debate on October 4th (Bolsonaro did not participate).  Across a variety of social media metrics, he has surged in the final days of the election.  Consensus expectations and traditional metrics, however, have Gomes coming in a distant third place.  At the same time, SMI of the consensus second place finisher, Haddad of the PT, has slumped.

This configuration of Gomes surging and Haddad slumping throws the consensus finish into doubt.

Traditional election metrics, such as polls and betting markets, have not picked up any major shifts for Gomes, who has remained stagnant over the last month with polls staying around 11%.  In fact, the CNT poll, which could be the final poll to be released (dated October 4-5), has him at 9.9%.  The same poll has Haddad at 24%.

Predictit, the betting market, has the probability of Gomes becoming president at 1%.  Predictit puts Haddad at a 33% probability.

This has created a highly unusual situation in which SMI has made a very late shift which has not been picked up in the slightest by traditional metrics.  In reality, the final TV debate being held only three days before voting exposed election analysis to this possibility.

The remaining outstanding issue regarding the Brazilian Election is how or if a very late surge in SMI of Gomes could impact results.  As already stated, the SMI shift was late in the campaign and became more acute due to the assumed victory of Gomes in the final debate.  SMI normally works well as a leading indicator but given the tight time window it is unclear to what extent voters will adjust.

Given the current situation, the safest prediction is that Gomes will outperform polls tomorrow whereas Haddad will underperform them.  Movement should be limited by the late nature of the shift, but it should be significant enough to modify votes.

First place is much easier and clearer to forecast.  SMI has been pointing to Bolsonaro for over a year and has not wavered.  SMI forecasts Bolsonaro to receive 42% of the valid votes, excluding blank and null votes.  This was an out-of-consensus forecast when it was made, but with the latest polls pushing Bolsonaro higher, the forecast no longer seems so aggressive.  The most optimistic poll estimates released in the last few days have him at 36%.

Summarizing:

  • SMI is more optimistic than the consensus concerning Bolsonaro, placing his expected results higher than polls, SMI clearly shows Bolsonaro as the first round winner,
  • Until recently, SMI had Haddad as a strong second place finisher, but a winning final debate performance by Ciro Gomes has seen his SMI surge in a potential late reversal, which puts the second place finish much more up for grabs,
  • SMI is a leading indicator and works well if voters are given enough time to adjust, as is the case in almost every election.  Given these shifts have occurred so late in the campaign (the TV debate was held just days before voting), their impact is not as clear,
  • Traditional metrics have not picked up on any improvement of the chances of Gomes, in contrast to SMI which shows a tremendous uptick for Gomes post-debate.

For a more detailed explanation, analysis, and forecast please request the 9 page ZettaCap analysis published today.