Mexico
The United States' largest trading partner runs one of the world's most asymmetric relationships: economy, security and family income all routed through a neighbor that keeps renegotiating the terms.
(the most single-customer trade profile of any large economy on earth)
(about 3.5% of GDP, and a major income source in the poorest states)
(no democracy had ever elected all its judges; Mexico elects the rest in 2028)
A note on framing. On this site, as in American politics, Mexico usually appears as a direction - where migrants come from, where fentanyl transits, where factories relocate. This page treats it as the subject: a country of 130 million with the world's most consequential asymmetric relationship. The through-line to watch is sovereignty under asymmetry. Mexico's economy is inside an American trade framework, its security policy operates under American pressure, and many of its households live partly on American wages. The question every section below returns to is what self-government means when the three main levers sit in a neighbor's hands.
Sheinbaum's tightrope
Claudia Sheinbaum took office in October 2024 with more formal power than any Mexican president in the democratic era. Her movement, Morena, and its allies hold a constitutional supermajority; the opposition is reduced to remnants; her approval was still 68% in El Financiero's August 2026 poll. She has spent that power on a doctrine her government calls the cool head: absorb Donald Trump's threats without public confrontation, concede what is concedable, and keep the relationship transactional. The concessions have been real: ten thousand troops to the border, water deliveries under the old treaty, and the transfer of more than ninety cartel figures into American custody. Some were men Washington had sought since the 1980s.
The results have also been real, which is why the doctrine holds. Border crossings fell to lows not seen in decades; the tariff blows that landed on other partners were repeatedly postponed or softened for Mexico, whose USMCA-compliant goods remain exempt from the broad US tariffs; and Washington's public praise of Sheinbaum has been conspicuous even as its pressure grew. But the tightrope has a built-in wobble: each concession proves the pressure works, which invites more pressure. The cool head buys time. What it has not yet bought is any restoration of the leverage Mexico conceded - and the sections below are a tour of how much that is.
The judicial experiment
In June 2025 Mexico began something no democracy had ever done: electing its entire federal judiciary by popular vote. The first round replaced the Supreme Court and about half of the 1,600 federal judges in one stroke; the rest follow in 2028, a year later than planned. The reform was Morena's answer to a judiciary it described - not always wrongly - as slow, elitist and captured by money. Turnout was 13%. Most voters faced ballots listing hundreds of names they had never heard. Printed slates of recommended candidates, circulated in Morena circles, almost exactly predicted the Supreme Court results, and the elected bench is broadly aligned with the movement that designed the process.
The consequence is a state with very few internal counterweights: one movement now holds the presidency, both chambers, most governorships and, its critics say, the courts. Whatever discipline exists is internal to Morena. The market response was the textbook one: Moody's cut its outlook to negative in 2024, naming the reform, and in May 2026 downgraded Mexico to its lowest investment grade, mainly over fiscal strain and the state oil company, Pemex. The larger cost compounds in the investment arithmetic covered below: contracts are only as strong as the courts that enforce them, and enforcement has become harder to predict. The open question, flagged as such: how elected judges behave under pressure, since none have existed at this scale; a year in, the record is too short to say.
The cartel war becomes a state question
The security story of 2024-26 is three wars stacked on top of each other. The first is inside the Sinaloa cartel. In 2024 the organization's co-founder was abducted by a rival heir's faction and delivered to American agents on a New Mexico airfield; the betrayal split the cartel into factions whose fighting has killed thousands and spread along the smuggling routes. The second is the state against the Jalisco New Generation Cartel (CJNG), and it produced the term's most consequential operation: in February 2026, Mexican forces killed the cartel's founder Nemesio Oseguera Cervantes - El Mencho - one of the world's most-wanted traffickers. The retaliation burned across twenty states: highway blockades, torched vehicles, clashes that killed some twenty-five National Guard members and about thirty gunmen. The epicenter was Jalisco, whose capital, Guadalajara, was to host World Cup games months later. The state held; the point had been made in both directions.
The third war is jurisdictional, and it is the one with constitutional stakes. Washington designated the major cartels as terrorist organizations in early 2025, flew surveillance drones over Mexican territory, struck alleged trafficking boats in the Caribbean and the Pacific, and - by American press accounts - prepared options for strikes inside Mexico. In April 2026 federal prosecutors in New York charged the sitting governor of Sinaloa and nine other current and former officials as cartel collaborators, putting on official record the charge that parts of the Mexican state work for the cartel. The governor denies it and has stepped aside; Mexico's government says Washington has not shown sufficient evidence. Sheinbaum has drawn one line publicly and repeatedly: cooperation, extradition, intelligence - yes; foreign military operations on Mexican soil - never. The line has held through September 2026. Every mechanism on this page - the trade leverage, the security pressure, the indictments - is a way of testing it.
The ledger on violence itself carries the usual measurement caveat. Official homicide counts have fallen sharply under Sheinbaum, from about 87 a day in September 2024 to about 41 in August 2026, and her government advertises the 53% drop. Analysts note that disappearances have kept rising: the official registry has passed 130,000 names, though a government review found later activity records for about 40,000 of those listed. They also note that the infighting moved killing into places counted less well, and that Mexican homicide statistics have always measured the state's counting capacity as much as its safety. Both facts fit: the war on the headline number is being won, and the war underneath it is not over.
An economy with one customer
Mexico sells more than four-fifths of its exports to one buyer. That single number explains both the country's industrial rise - no other large economy sits inside American supply chains so deeply - and its negotiating position, which is that it does not have one. The framework holding the arrangement together, the USMCA trade agreement, came up for its scheduled review in 2026, and Washington used the calendar as leverage. It opened bilateral talks with Mexico in March, and on the July 1 decision date declined to extend the agreement's term; its talks with Canada collapsed in August into a tariff fight. The treaty did not end; it changed character. What had been a sixteen-year stability anchor became a rolling annual review running toward its 2036 sunset - which converts every year of Mexican policy into a renewal application, reviewed in Washington.
The macro numbers show what the uncertainty costs. Growth was about 0.5% in 2025, recession-adjacent, and 1.1% in the first half of 2026 - slow for a country that needs roughly triple that to absorb its labor force. The nearshoring boom that was supposed to follow the China decoupling has been real in announcements and modest in construction. Analysts' shorthand - nearshoring without growth - captures investment held back less by tariffs than by uncertainty itself, judicial and trade alike. Remittances, the income floor of the rural south, fell in 2025 for the first time in eleven years; analysts at the bank BBVA blame fewer new migrants joining the US workforce, a stronger peso and the fear US immigration raids spread. They have since steadied, rising 3.1% in January-July 2026 despite a 1% US tax on cash-funded transfers from January. And Pemex still carries $77.5 billion of financial debt (June 2026, down 9% this year), a standing claim on public money that every budget must service before it builds anything.
Against this stands the ledger Morena runs on, and it is not propaganda. Poverty fell by more than ten million people between 2018 and 2024 on the official measure, and the minimum wage has risen about 150% in real terms since 2018, by the government's count. The wage and social policies behind those numbers are widely credited with renewing the movement's majorities. The model's constraint is arithmetic, not ideology: transfers funded by a slow-growing economy, under a trade framework now reviewed every year with a foreign government, with the peso and the bond market watching the courts. It works if growth returns, and strains if it does not.
The paths from here
Dated September 2026, with US-Mexico talks aiming for a trade deal before the American midterms in November and the first annual USMCA review due in 2027.
Managed asymmetry
The current equilibrium extends: annual trade reviews traded for security concessions, the cool head absorbing each new demand, no rupture and no restoration. Mexico remains prosperous enough to stay stable and constrained enough that the prosperity stays modest.
Will it happen? It is the base case because both governments profit from it: Washington gets compliance without the costs of acting alone, Sheinbaum gets results she can show voters. The continuing trade talks fit it. Its weakness is that it depends on personalities - a Mexican president this disciplined and an American one this transactional - and both offices change hands within the decade.
The strike, and the rupture
Washington acts unilaterally on Mexican soil - a drone strike on a lab, a raid on a capo - whether or not cartel violence provides the trigger. Mexican politics does the only thing its history allows: cooperation collapses, the cool head becomes politically impossible, and years of carefully built security integration unwind in months.
Will it happen? The capabilities are deployed, the legal designations are in place, and by press accounts American planners have prepared the option. Against it: most analyses, American ones included, conclude a strike would end the cooperation that produces the actual results, and none had happened by late September 2026. This is the tail risk both presidents manage around daily - and the single most consequential branch on this page.
The institutional slide
The fused state hardens: elected courts defer, indictments of officials multiply, investment stays away, and Mexico settles into a dominant-movement system - elections held and won, checks internal to the movement, growth permanently below need. Not collapse; entrenchment.
Will it happen? Much of the structure is already in place, and the rating agencies are wary - Moody's downgraded Mexico in May 2026 and S&P's outlook is negative - though mainly over fiscal strain. What could interrupt it: a Morena succession fight (the movement has handed on the presidency only once, from its founder), an economic shock that breaks the transfer arithmetic, or elected judges ruling against the government - the experiment with little data yet.
The bet pays
Uncertainty clears enough - a stabilized trade framework, a post-Mencho security plateau - that the deferred nearshoring pipeline converts to construction. Growth returns toward 3% late in the decade, and Mexico finally collects the dividend its geography has promised since 1994.
Will it happen? The ingredients are present: wages competitive with China, the deepest American supply-chain integration on earth, an energy and water build-out in the north actually underway. Every input except one - predictability - favors it. That one input is priced by the courts and the annual review, and neither has yet settled.
Fragmentation's second act
The Sinaloa split and El Mencho's death do to organized crime what decapitation has so often done: multiply it. Smaller, hungrier factions diversify into extortion of the legal economy - avocados, limes, ports, construction - and violence shifts from cartel-on-cartel to cartel-on-everyone, hollowing the formal economy state by state.
Will it happen? It is the usual lesson of kingpin removals since the 1990s, but the early post-Mencho record is mixed: official homicides kept falling, while extortion barely moved (down 2.8% in January-August 2026, officially). The counterweight is that extortion, unlike trafficking, attacks voters directly - it creates political demand for state presence that drug smuggling never did. Which force wins decides whether the security story of the 2030s is better or much worse.
Where serious analysts disagree
Whether American pressure is working. The case that it works points at outcomes: record-low crossings, falling fentanyl seizures read as falling shipments, unprecedented prisoner transfers, a cartel founder dead. The case that it corrodes points at what produced them: a neighbor's sovereignty converted into an annual performance review, cooperation that depends on humiliation staying private, and a Mexican public whose nationalism is the one force no Mexican government can defy indefinitely. Both cases describe something real, which is the problem: the method's results and its risks come from the same place.
Whether "narco-state" is description or slur. The American indictment of a sitting governor gave the strong version official weight: in parts of the country, the boundary between crime and state has dissolved, and pretending otherwise is diplomacy. The counter-reading, common among Mexico specialists, is that capture is real, regional and contested, not national: the federal state fights cartels, loses ground here, wins there, and a label that flattens Culiacán and Mexico City into one word explains neither. Where a reader lands matters practically, because the first reading is used to justify foreign operations and the second argues against them.
Whether the judicial election deepens democracy or ends its separation of powers. Its defenders are not wrong that the old judiciary served money and impunity comfortably, or that elite disgust with the reform tracks elite comfort with what preceded it. Its critics are not wrong that a movement electing the judges who will judge the movement is the oldest trick in the dominant-party handbook. The evidence that would settle it - how the elected bench rules when Morena itself is the defendant - is still thin a year on. The court's president says it follows no agenda but the constitution's; critics point to rulings that narrowed protection against government acts.
The mechanics behind this
Strip the specifics and this page is four site-wide mechanisms in one country. Asymmetric interdependence: integration you cannot exit becomes leverage for whoever controls the door. Decapitation dynamics: removing network hubs fragments networks rather than ending them. Dominant-movement politics: when checks move inside the ruling coalition, the constraint on power becomes the coalition's own succession problem. And measurement politics: homicide counts, seizure statistics and poverty lines each measure the counter as much as the counted. The country holding all four at once, next to the superpower running its own stress test, is why this page and the United States page are best read as a pair.