The El Financiero poll published this week says the current presidential race is Sheinbaum 48, Gálvez 32. That is a 16 point gap, which is enormous, but is also closer than any other poll. The Oraculus poll of polls database shows most polls give Sheinbaum a 20 to 30 point lead.
So the El Financiero poll is a bit of an outlier, but let's take it as accurate just to work through a best-case scenario for Mexico's opposition. What does this poll say about the state of the race?
Gálvez’s approval is rising, but still net-negative. The opposition candidate has 37% approval and 44% disapproval. As more people have learned about Gálvez, her approval rating has gone up. However, a negative campaign by the AMLO government and some strong party loyalty from Morena voters have also increased her disapproval rating. That said, I read these numbers as saying Galvez is someone who can improve her standing in the coming months, particularly as she gets more air time in the media and appears in the debates.
Sheinbaum's approval rating is steady. Gálvez’s improvements aren't coming at a cost to the government candidate. Sheinbaum's has 49% approval and 33% disapproval and that number has held relatively steady for months. Sheinbaum is smart, cautious, and unlikely to make mistakes that cost her significant support.
Morena leads the Congressional vote 41-31 over the PAN-PRI-PRD coalition, with MC taking 11 percent. Morena won 44-41 in the Congressional midterms of 2021 and that was enough to take a majority in the Congress. A ten point gap would be a more substantial margin of victory (and coalition partners would add to that margin). The 11 percent for MC, while not making much of a dent in the presidential race right now, would be the party's best showing in Congress.
Few undecideds. The El Financiero poll says 10 percent of voters are undecided and around 18 percent of voters have a preference but could change their mind. Nearly all the polls in the country suggest the number of undecided voters is under 20%.
Given the numbers above, there are two ways for Xochitl Gálvez to win this race. One is to chip away at Sheinbaum's lead month after month until she is running near even. While I think there is room for the opposition to gain further ground, there aren't enough undecided and moveable voters for this strategy to hit the tipping point by July. Gálvez would need to win nearly all the undecideds, tip a large portion of MC votes her way, and convince some Sheinbaum voters to flip. The other way to shake this up is for some scandal or campaign moment to cause a giant shift in momentum in the race. In World Politics Review last week, I covered the recent scandals around AMLO. While polling suggests those scandals have made a small dent in AMLO's otherwise rock-solid approval rating, I would stress it's a small effect. And if corruption by AMLO's brothers and children isn't moving the electorate in a substantial way, it's hard to imagine another scandal that will.
That's why the presidential campaign is currently Claudia Sheinbaum's race to lose. Even if you take Gálvez’s best poll of the campaign, the numbers just aren't there for her to gain the ground she needs. And yet, it's worth keeping it close for the opposition. Keeping the Congress close and limiting the margin of victory will matter for Sheinbaum's mandate and the opposition's chances at future electoral gains. And electoral surprises do happen, but only if oppositions are prepared to take advantage when the conditions strike.
One other interesting polling note: A column in Milenio this week highlighted the fact that Mexico's recent national survey about insecurity in the country was unable to poll in some regions due to.... insecurity. That is a clear methodology issue. No matter AMLO's insistence that security has improved in the country, the situation is not better than when he took office six years ago and is arguably worse in some respects. And if the government can't collect the data it needs about insecurity, it's going to be that much more difficult to measure and improve.