On April 1, 2024, Bertha Gisela Gaytán stood at a campaign rally in Celaya, a city in Guanajuato that’s been torn apart by a cartel turf war. She was running for mayor, promising to fix the roads and get the violence under control. Weeks before, she had asked the federal government for protection. She never heard back. That afternoon, gunmen opened fire on the rally and killed her, along with a city council candidate standing next to her. She never got to finish the campaign.
What happened to Gaytán isn’t some rare, tragic exception. It happens over and over, in towns most people outside Mexico have never heard of.
In Mexico, you can vote for mayor. You just can’t vote against the cartel’s pick. In large parts of the country, organized crime has quietly taken over the job of deciding who’s allowed to run for local office, who gets pressured into quitting, and who gets killed for refusing. On paper, Mexico looks democratic. On the ground, in a lot of towns, it isn’t.
You wouldn’t necessarily know this from the news. Mexico just elected its first female president. Turnout was high. The courts kept functioning. Nationally, everything looked fine. But once you zoom into the municipal level, that picture falls apart pretty fast.
Why This Should Worry You
Democracy isn’t only about whether people get to vote. It’s also about who they’re allowed to vote for. If the list of candidates has already been narrowed down by threats and assassinations before election day, the vote itself doesn’t mean much anymore. A country can hold clean elections, have working courts, and still not really be a democracy if criminal groups get to decide who’s allowed to lead.
That’s more or less where Mexico is right now, even if it doesn’t show up in the national numbers.
But the National Election Went Fine
Someone could reasonably push back here: Mexico just had a historic, mostly peaceful presidential election. Isn’t that what actually matters?
I don’t think so, or at least not entirely. Most of what actually affects people’s day-to-day lives gets decided locally, not nationally. A smooth presidential race doesn’t undo a mayoral race that was decided by a death threat months earlier. Gaytán wasn’t killed in some remote, forgotten village. She was killed in the middle of a real campaign, after asking for protection that never came. The national system can work exactly as intended while hundreds of local ones quietly don’t. That gap is kind of the whole point.
The Violence Isn’t Random. It’s Scheduled.
Cartel violence against politicians isn’t spread evenly across time. It follows the election calendar almost exactly. Data Cívica’s project https://votar-entre-balas.datacivica.org and ACLED Press Release https://acleddata.com/press/violence-during-2024-elections-exceeded-levels-recorded-during-2018-and-2021-elections-press have been tracking this for years, and the pattern holds up: violence spikes right before elections and basically disappears in between. 2026, a non-election year, saw almost no candidate assassinations. 2024, Mexico’s biggest election cycle in recent memory, was the deadliest on record, with 130 violent attacks on political figures. Of those victims, 34 were murdered, 40 survived attacks, 32 received threats,10 candidates were kidnapped, and 14 armed attacks during which the candidate survived but someone else died.
And it wasn’t presidential candidates getting targeted. More than 80% of the attacks hit local races — mayors, city council members, local police chiefs. Cartels don’t really need the presidency. They need control over highways, ports, and municipal budgets, and that kind of control runs through city hall, not the national palace.
To get it, they use a doctrine that’s become almost a slogan at this point: plata o plomo. Silver or lead. Take the bribe or take the bullet. In 2024 alone, at least 130 candidates were physically attacked or kidnapped. Over 550 — Gaytán included — asked for police protection just to keep campaigning. Hundreds more dropped out quietly, no public explanation given, because everyone already understood why. By the time voters actually show up to the polls, the real filtering has already happened.
Cartels Don’t Just Corrupt the State. They Replace It.
In parts of Guerrero, Michoacán, and Chiapas, cartels aren’t just influencing local government from the shadows — they’re functioning as local government. They collect informal taxes from businesses, often more reliably than the actual tax office does. They settle disputes that never make it to a courtroom. They pay for road repairs, fund town festivals, hand out the kind of favors a mayor’s office used to hand out. None of that comes with an election. The violence around campaign season isn’t random chaos on top of that system — it’s how that system protects itself.
In political research published by political scientist Andreas Schedler https://journalofdemocracy.org/articles/the-criminal-subversion-of-mexican-democracy, he calls this “criminal subversion.” His point, roughly, is that this kind of violence doesn’t destroy elections so much as it captures them from the inside. It lines up with something political theorist Guillermo O’Donnell called “brown areas” — pockets of a country where the state’s laws technically exist but don’t really apply in practice. In Mexico’s brown areas, people still go vote. It just doesn’t change much, because the choice they’re making was narrowed down long before they got to the ballot box.
What Fixing This Actually Requires
Sending soldiers to guard polling stations on election day is dealing with the symptom, not the actual problem. By then, the candidates who mattered have already been threatened, bought off, or killed. The real fight happens months earlier, during the part of the campaign season nobody’s paying close attention to — the part Gaytán didn’t survive. Fixing this would mean going after the financial networks that let cartels run towns in the first place, and giving the state, not armed groups, the actual authority over who’s allowed to run for office. It would also mean something a lot more basic: answering a protection request before the rally, not after the funeral.
Until something like that happens, Mexican elections are going to keep producing winners who weren’t really chosen by voters. The ballots themselves are legal. The outcome, in a lot of towns, isn’t. Somewhere between the campaign trail and the ballot box, the actual election already happened, and most people never got a say in it.
Cartels use fearmongering to retain control of a government. It is very fair for you to speak out and tell the people that letting people like this control your government will hold it back. But how can we reduce their fear? Not listening to the Cartel or not letting them do what they want is seen as a death sentence, so there must be protections given to the brave people who want change. Help is needed from the outside so the people of Mexico can live in their homes without fear.
Considering I’ve looked at a fair amount of case studies regarding democratic erosion in Mexico, I am completely shocked that I have not heard of this, at least in the modern day. I was especially shocked by the statistic on the 130 candidates attacked or kidnapped. I completely agree that the nation cannot move forward when so much of the government is manipulated by these cartels, but I also agree with the previous commenter that extensive measures need to be taken to protect the people willing to stand up against the cartel. The cartel has proven time and time again that they are going to go to extreme measures to prevent candidates they don’t want from having power. I agree the financial channels that feed them need to be choked but that will take time. How will people be protected in the meantime?
This post is incredibly interesting, and I am additionally shocked that I was not aware of this matter. I found your “why this should worry you” section to be very interesting as well. Democratic participation can be thought of generally by most, as people having the right to vote. But, using this lens, what happens when violence stands in the way of who voters can actually vote for? These acts of violence are undermining democratic participation for sure.
Great article. I found it particularly interesting to learn about the impact of cartels on local election outcomes, as it shows how democracy can be undermined even when elections are being held. It is hard to grasp that even with dominant government institutions in place, democracy can still be breached. This made me think about my experience living in Hawaiʻi, where colonial amendments dictated local elections for a long time and determined who governed our land. This makes me wonder: could an entity like a cartel ever become so powerful in the US that it has the same effect?
I thought your distinction between national and local democracy was particularly interesting. The argument that a country can have a relatively successful national election while democratic competition is being undermined at the local level is a perspective I had not really considered before. Your discussion of how cartels can influence who is even able to appear on the ballot also makes the concept of democratic erosion feel much more concrete. I especially liked the connection to Schedler’s idea of “criminal subversion,” because it shifts the focus from whether elections technically occur to whether voters actually have meaningful choices.
As someone who is into politics and also researches true crime, I found this post really intriguing. I thought your point about cartel violence being related to elections was especially interesting because not only did I used to think cartel involvement was primarily drug-related, but it shows that these attacks are not just random acts of violence. It made me think differently about what actually counts as a representative leader. Even if there are many voters, they may not have a real choice if candidates drop out because of cartel threats. I wonder what could be done earlier in the process to protect candidates so they don’t succumb to fear.
I enjoyed your point on that elections can appear democratic nationally while being undermined at the local level. I think that even if voting is free, democracy depends on voters having meaningful choices. If violence and/or intimidation is what determines who can run for office, electoral competition is weakened before voters can make a decision.
This is a really important point because it shows that democracy has to be measured by more than voter turnout or whether elections formally take place. If candidates are threatened, forced to withdraw, or killed before voters ever reach the polls, then political competition has already been distorted. What I found especially striking is the distinction between national and local democracy: Mexico can hold a largely functioning presidential election while some municipalities experience a very different political reality. The discussion of “criminal subversion” also helps explain why cartel violence is not simply ordinary crime—it can directly shape who is able to participate in politics and, ultimately, what choices voters actually have. The case of Bertha Gisela Gaytán makes the issue particularly powerful because it demonstrates how failures to protect candidates can undermine democratic participation long before election day.
Your point that the real election takes place before anyone votes is convincing. The evidence you cite makes it even sharper: violence follows the electoral calendar and then fades, yet those same groups govern parts of the country from the shadows all year. Campaign-season violence, then, may signal a struggle over control rather than a functioning system under pressure. Once a group has secured a town, it has little reason to challenge the candidates. If that pattern holds, disputes should gather in places where control remains contested—especially during turf fights—and appear less often where one group has already secured dominance. That would complicate the claim that elections are simply captured: in towns under the firmest control, voting may look calm precisely because genuine competition has disappeared. Answering protection requests before the rally helps most where control is disputed, but it does little in towns where candidates quietly drop out and no attack ever appears in the data. My reading predicts that counts like the 130 attacks you cite understate capture, and that the places with the fewest attacks include some of the most captured municipalities.
This blog post was really well structured and effectively organized. It called into question the idea that governments can offer freedoms that citizens are able to exercise, but in a lot of ways feel as though they cannot, for fear of their safety. The solution you offered for this issue did a great job of conveying that while a solution is possible, it will be difficult, and the government needs to answer this call to action before the issue is already over. The evidence you chose to support your topic furthers the fact that this isn’t an isolated incident, but instead a targeted attack that happens more often than people choose to recognize.
I found your post to be extremely insightful, and I am shocked by your findings. The point that you make about political violence targeting local rather than national elections is very interesting and raises a large concern about the foundation of democracy in Mexico. An important aspect of democracy is that citizens should be able to formulate their preferences, signify their preferences to their fellow citizens and government through individual and collective actions, and have their preferences weighted equally in the decision making process. If cartels are specifically scheduling their political attacks and not allowing voters to participate in government on their own accord, it is really limiting the legitimacy of democracy which is very frightening. Additionally, this coincides with your questioning of how real the election results are. The evidence you used is also very interesting and backs up your points well, especially the statistics on political violence during election years versus non-election years as it showcases how these are not one off instances.