MX-I-02: The Mexican Armed Forces and the Militarisation of Public Life (2006–2026)

Status: DRAFTWords: 10,507

1. Key Takeaways

  • Mexico's twentieth-century civil-military settlement was, in comparative Latin American terms, an anomaly of success: an army born of the Revolution was progressively subordinated to civilian — meaning PRI — authority without a single successful coup after 1920's institutionalisation, through a sequence usually marked by Cárdenas's 1938 incorporation of the military as a party sector, Ávila Camacho's 1940 removal of that sector from the party, and the 1946 inauguration of Miguel Alemán, the first civilian president of the post-revolutionary era, after which no serving general ever again held or seriously contested the presidency. The price of subordination was a barracks bargain: the armed forces accepted political abstention and modest budgets in exchange for near-total institutional autonomy — a serving general as Secretary of National Defence (SEDENA) and a serving admiral at SEMAR in every cabinet to the present day, self-managed promotion and justice systems (the fuero militar), and an opacity that exempted the institution from the scrutiny applied to the civilian state. Every development treated in this document is a renegotiation of that bargain.

  • The bargain contained two standing exceptions that prefigured the post-2006 era. The first was doctrinal: the armed forces' mission set — DN-I (external defence), DN-II (internal security), and DN-III-E (the disaster-relief plan instituted in 1966 that became the institution's principal source of public affection) — always included an internal-security mandate that civilian governments could activate, and did, from railway strikes to the 1968 Tlatelolco massacre. The second was the dirty war: the army's counter-insurgency campaign in Guerrero and elsewhere across the late 1960s–1980s, with enforced disappearances, clandestine detention, and the documented practice of death flights, was never judicially accounted for — the Fox-era FEMOSPP special prosecutor's office (2002–2006) collapsed without convictions, and the AMLO-era truth commission on the 1965–1990 period reported persistent SEDENA obstruction of its archives [TBD-VERIFY: the commission's 2023–2024 findings on archive access and the scale of documented disappearances]. The impunity precedent — that the army performs internal missions and is not held to account for them — is the institutional inheritance the drug-war deployments activated at scale.

  • Felipe Calderón's 11 December 2006 Operativo Conjunto Michoacán (the drug-war narrative is carried at MX-J-02 and MX-A-02; this document treats the institutional consequences) converted an exceptional internal-security mandate into a permanent constabulary role. By the deployment peak around 2010–2011, on the order of 45,000–50,000 troops were assigned to policing tasks [TBD-VERIFY: peak deployment figures vary by source between roughly 45,000 and 96,000 depending on whether support roles are counted]; CNDH complaints against SEDENA rose roughly tenfold from pre-2006 levels [TBD-VERIFY: CNDH annual complaint statistics — commonly cited figures show SEDENA complaints rising from a few hundred annually before 2007 to over 1,600–1,800 at the 2009–2012 peak]; and the army leadership itself repeatedly and publicly requested a legal framework for missions it was performing without one — a constitutionally anomalous situation that persisted for thirteen years, since Article 129 of the Constitution provides that in peacetime no military authority may exercise functions other than those directly connected with military discipline.

  • The legal architecture of the deployments rested for two decades on a jurisprudential workaround rather than a constitutional mandate: the SCJN's 1996 resolution of acción de inconstitucionalidad 1/96, which held that the armed forces could lawfully act in support of civilian authorities, at their express request, without a declaration of war. Attempts to regularise the situation failed twice before succeeding on the third, most militarising attempt: the December 2017 Ley de Seguridad Interior (Peña Nieto) was struck down in its entirety by the SCJN in November 2018; the March 2019 constitutional amendment creating the Guardia Nacional (GN) constitutionalised a nominally civilian force while a transitory article authorised the president to use the armed forces in public security for five years (to March 2024) under the Inter-American Court's "extraordinary, regulated, subordinated and complementary" standard; that window was extended to 2028 by constitutional amendment in November 2022 [TBD-VERIFY: the extension's exact promulgation date]; and in September 2024 the outgoing AMLO supermajority amended the Constitution to place the GN permanently under SEDENA [TBD-VERIFY: promulgation date and final vote], reversing by constitutional means the SCJN's April 2023 ruling that had struck the same transfer when attempted by ordinary law.

  • The central paradox of the period — flagged at MX-J-02 §7 and developed here institutionally — is that Andrés Manuel López Obrador, who campaigned on returning the army to its barracks and whose abrazos no balazos doctrine promised demilitarisation, presided over the largest expansion of military functions in modern Mexican history. The GN was built to over 100,000 members, overwhelmingly recruited and seconded from the army and military police, commanded in practice by serving or retired generals [TBD-VERIFY: force size circa 118,000–130,000 by 2023–2025 and the share of military origin, commonly reported around 75–85%]. Beyond security, the armed forces became the administration's implementing agency of choice: the construction and operation of the Felipe Ángeles airport (AIFA) and other airports; the Tren Maya's construction and its operating company; the takeover of customs (2021) and ports and merchant marine (2020); the revived Mexicana de Aviación airline under a SEDENA company (2023); the construction of thousands of Banco del Bienestar branches; pipeline surveillance against huachicoleo; and COVID-19 vaccine logistics. By the end of the sexenio the military was a construction conglomerate, airport operator, airline owner, customs collector, and tourist-railway operator — with revenue streams and trusts of its own.

  • Three readings of AMLO's choice circulate at strength and are not mutually exclusive. The trust reading takes AMLO at his word: he described the armed forces as pueblo uniformado — the people in uniform — and as one of the only national institutions not rotted by neoliberal-era corruption, hence the safe custodian of his projects. The delivery reading is managerial: the army builds on time, does not strike, is exempt in practice from procurement law and transparency obstacles, and could be commanded directly — the bureaucracy of last resort for a president distrustful of the civil service he inherited. The loyalty-purchase reading is political: by giving the institution missions, budgets, businesses, and constitutional permanence, AMLO bound the armed forces to the 4T project and insured his transformation against the institution best placed to obstruct it. The corpus holds all three; the accountability record in §5 is the evidence against which each must be tested.

  • The accountability record of the expansion is thin and in key episodes openly adverse. The army's role in the Tlatlaya killings (June 2014) and in the Ayotzinapa disappearance (September 2014; MX-B-03, MX-J-01) produced, respectively, collapsed prosecutions and a decade of documented SEDENA resistance to disclosing its own records — resistance that the AMLO-era truth commission (COVAJ) and the special prosecutor confronted directly in 2022, when the prosecution of military suspects was followed by the resignation of special prosecutor Omar Gómez Trejo amid charges that the FGR had undermined his cases, and the GIEI international experts departed in 2023 naming SEDENA's stonewalling [TBD-VERIFY: precise sequence of the 2022 warrant withdrawals and the GIEI's final-report findings]. The September 2022 Guacamaya hack of SEDENA's servers — several terabytes of internal email — revealed military surveillance of journalists, activists, and political actors, the army's continued use of Pegasus-type spyware after AMLO's pledge that espionage had ended, military intelligence monitoring of the Ayotzinapa investigations, and undisclosed details of the president's health [TBD-VERIFY: specific revelations and their documentary basis]. No comparable institution in the Mexican state combines this functional weight with this little external oversight: defence committees in Congress are weak, the GN's nominal civilian organs lack purchase, and national-security classifications — including the November 2021 presidential accord declaring the sexenio's megaprojects matters of national security — shield military-run projects from freedom-of-information law [TBD-VERIFY: the accord's litigation outcome before the SCJN].

  • The October 2020 arrest of General Salvador Cienfuegos — Defence Secretary for the entire Peña Nieto sexenio — by US authorities at Los Angeles airport on DEA-built trafficking-protection charges, and the extraordinary sequence that followed (Mexican pressure including threats to restrict US agents; the US dropping all charges in November 2020 and repatriating him; the FGR exonerating him within weeks; AMLO publishing the DEA case file and legislating new restrictions on foreign agents in December 2020), is the single most revealing episode of the period for what it showed about the bargain's terms: the Mexican state, across its civilian leadership, treated the protection of the army's institutional honour as a core national interest worth a near-rupture with Washington — and the army registered the lesson that it would be protected [TBD-VERIFY: exact sequence and the contents of the published file]. The episode crippled US–Mexico security cooperation at the institutional level for years and prefigures the sovereignty politics of the 2025–2026 FTO-designation era (MX-E-02).

  • The Sheinbaum administration (from October 2024; MX-D-03) inherited the consolidation as a fait accompli — the GN constitutionally under SEDENA, the military enterprises operating — and has not reversed any element of it; but it has built alongside it a civilian counter-model: Security Secretary Omar García Harfuch's intelligence-led, civilian-commanded operational strategy, backed by 2024–2025 legislation restoring investigative and intelligence capacity to the civilian security ministry. The result is a dual structure whose internal balance is the central institutional question of the sexenio — sharpened externally by the second Trump administration's designation of cartels as Foreign Terrorist Organizations (February 2025) and its escalating discourse of unilateral US military action against cartel targets, against which Sheinbaum has drawn explicit sovereignty red lines ("coordination yes, subordination no") while quietly recalibrating bilateral military-to-military cooperation [TBD-VERIFY: 2025–2026 bilateral security-framework developments and reported US strike authorities].

  • Comparatively, Mexico is the largest case in a hemispheric militarisation wave — Brazil's GLO deployments and Bolsonaro-era military colonisation of civilian office, El Salvador's Bukele model as the region's pole, the Guatemalan, Honduran, and Ecuadorian deployments — but it is distinct in kind: the Mexican military has no coup tradition, has never contested an electoral outcome, and has expanded by civilian invitation, sexenio after sexenio, across parties. The literature's formulation — militarisation without militarism: a vast functional expansion without praetorian intent — captures the distinction; its critics reply that the distinction decays, because an institution that acquires budgets, businesses, and constitutional missions acquires interests, and interests acquire veto power whether or not anyone plans a coup. Whether the 2006–2026 expansion proves a pendulum (a reversible emergency posture, as the old pact's defenders hope) or a ratchet (a new military-state equilibrium) is the question on which this document closes.

2. The Old Pact and Its End (1946–2006)

2.1 The Demilitarisation of the Presidency and the Barracks Bargain

The Mexican Revolution produced a state ruled by its generals — every president from Carranza through Ávila Camacho was a revolutionary military man — and the central achievement of post-revolutionary institution-building was the conversion of that army of caudillos into a subordinate state institution. The standard sequence runs: Calles's 1920s professionalisation campaigns under Joaquín Amaro, which rotated zone commanders, shrank the officer corps, and broke the personal armies; Cárdenas's 1938 incorporation of the military as one of the four sectors of the corporatist party (the PRM), a move conventionally read as paradoxical subordination — bringing the soldiers into politics in order to dilute them among the peasants, workers, and popular sectors; Ávila Camacho's 1940 removal of the military sector from the party, ending the armed forces' corporate political representation; and the 1946 inauguration of Miguel Alemán Valdés, the first civilian president of the post-revolutionary era. After 1946, no serving general held the presidency again, and after the 1952 candidacy of General Miguel Henríquez Guzmán — whose dissident campaign and its repression marked the last military-political challenge from within the revolutionary family — none seriously contested it.

What stabilised the settlement was not civilian control in the Huntingtonian sense — no civilian defence ministry, no parliamentary scrutiny, no civilian penetration of military affairs — but a bargain. The armed forces accepted political abstention, conspicuous loyalty to the presidency, and budgets that were, by Latin American standards, remarkably small (Mexican defence spending ran well under 1% of GDP for most of the period, among the hemisphere's lowest). In exchange they received institutional self-government: the Secretary of National Defence has been a serving army general, and the Secretary of the Navy a serving admiral, in every cabinet from the 1940s to the present — Mexico has never had a civilian defence minister; promotions, doctrine, education, and discipline were managed internally; military justice (the fuero militar, Article 13 of the Constitution and Article 57 of the Code of Military Justice) kept soldiers' conduct, including conduct against civilians, inside military courts; and the institution was exempted in practice from the audit and transparency regimes applied to the civilian state. Scholars of Mexican civil-military relations — Roderic Ai Camp's Generals in the Palacio (1992) and Mónica Serrano's work on the pact are standard references — emphasise that the arrangement was less "civilian control" than mutual non-interference: the civilians did not look inside the barracks, and the army did not look outside them.

2.2 The Mission Set: DN-I, DN-II, DN-III

The bargain's terms were visible in the armed forces' own doctrinal triad. DN-I covered external defence — the nominal core mission, in practice residual for a country whose only land-war scenario involved an unfightable neighbour. DN-II covered internal security — the assistance of civilian authority in maintaining internal order, the legal-doctrinal door through which every domestic deployment passed. DN-III-E, the disaster-relief plan instituted in 1966 [TBD-VERIFY: standard accounts date Plan DN-III-E to 1966, formulated after the 1966 Pánuco river flooding], became the institution's principal interface with the public and the foundation of its standing atop institutional-confidence surveys: the army that dug survivors from the 1985 Mexico City earthquake rubble (a performance in fact heavily criticised at the time), evacuated flood zones, and ran vaccination posts was the army most Mexicans encountered. The triad mattered analytically because it normalised, inside military doctrine and public expectation alike, the premise that the army's domestic employment was ordinary — a premise in permanent tension with the constitutional text examined in §3.

2.3 The Dirty-War Exception and the Impunity Inheritance

The pact's dark side was the use of the army as the regime's instrument of last resort against internal enemies, with guaranteed impunity. The army broke the 1958–59 railway strikes; it occupied the National University in 1968; soldiers were central to the 2 October 1968 Tlatelolco massacre, and the army provided the institutional context of the 10 June 1971 Halconazo paramilitary attack. Most systematically, the counter-insurgency campaigns against the rural guerrillas of Guerrero — Lucio Cabañas's Partido de los Pobres and Genaro Vázquez's ACNR — and against the urban Liga Comunista 23 de Septiembre across the late 1960s–1980s involved enforced disappearance as method: clandestine detention in military installations (the Pie de la Curva barracks in Acapulco and Military Camp No. 1 in Mexico City recur in survivor testimony), torture, and the documented practice of vuelos de la muerte — death flights dropping prisoners into the Pacific [TBD-VERIFY: scale estimates; registers associated with the truth-commission process document several hundred disappeared in Guerrero alone, with national dirty-war disappearance estimates commonly given in the high hundreds to low thousands].

The accounting never came. The Fox-era Fiscalía Especial para Movimientos Sociales y Políticos del Pasado (FEMOSPP, 2002–2006) — the democratic transition's designated reckoning — produced a draft historical report (leaked in 2006 after official editing disputes), indicted former president Echeverría for genocide in proceedings that ultimately collapsed, and closed without a single final conviction of a military perpetrator. The AMLO-era Comisión para el Acceso a la Verdad covering 1965–1990 (created October 2021) gained formal access to military installations and archives but reported, in its 2023–2024 outputs, persistent SEDENA obstruction — documents withheld, archives reorganised or reported lost, site visits managed [TBD-VERIFY: the commission's specific findings, the internal disputes over its final report in 2024, and the documented instances of archive obstruction]. The relevance to this document is direct: the dirty war established, and its non-accounting confirmed, the institutional precedent that the army performs internal missions at presidential direction and answers for them to no external authority. When the Ayotzinapa investigations of the 2010s–2020s ran into SEDENA's archival wall (§5.2), they were encountering a sixty-year-old institutional reflex.

2.4 The 1990s Creep and the December 2006 Rupture

The pact began visibly straining before Calderón. The January 1994 Zapatista uprising put the army into a counter-insurgency theatre under media scrutiny the dirty war never faced. The 1990s drug-mission creep brought the institution's first great public corruption scandal: the February 1997 arrest of General Jesús Gutiérrez Rebollo, the serving drug czar, for protection of the Juárez cartel — an early demonstration that proximity to the trade corrupted the military as it had the police. Zedillo and Fox both expanded military participation in counter-narcotics, and military officers were seconded throughout the civilian police apparatus.

The rupture nonetheless belongs to 11 December 2006. The Operativo Conjunto Michoacán and the deployments that followed (the operational narrative at MX-J-02 §3; the presidency at MX-A-02) changed the institution's situation in four ways that outlasted every subsequent strategy debate. First, scale and permanence: tens of thousands of troops in continuous constabulary deployment, year over year, converted DN-II from exception to standing mission; the army reorganised training, basing, and doctrine around a policing war it had not sought. Second, exposure: the CNDH complaint surge, the emblematic abuse cases, and the international human-rights litigation of §3.4 ended the institution's inviolability to external scrutiny precisely as its public role expanded. Third, leverage: a state that could not fight its war without the army owed the army — and the army's leadership, notably Secretary Guillermo Galván under Calderón and Salvador Cienfuegos under Peña Nieto, openly invoiced the debt, repeatedly stating that the armed forces were performing missions without a legal framework and requesting one, alongside budget growth, equipment, and expanded fuero protections. Fourth, indispensability as policy trap: each sexenio entered office noting that the deployments had failed to reduce violence and that police reform was the real answer, and each, finding the police unreformed, extended the deployments. The institutional consequence of December 2006 was thus not simply that the army went into the streets but that the political class lost, within a decade, any practical concept of public security without it.

3.1 The Article 129 Problem

The constitutional text is unusually explicit. Article 129: "In time of peace, no military authority may exercise functions other than those that are directly connected with military discipline." The provision is an 1857 inheritance, written by liberals against the praetorianism of the Santa Anna era, and it sits alongside Article 89, fraction VI, which empowers the president to dispose of the permanent armed forces "for internal security and the external defence of the Federation." The two-decade constitutional controversy over the deployments is, at base, the contest between these two articles: critics read Article 129 as a flat peacetime prohibition on military performance of civilian public-security functions; governments read Article 89-VI as an autonomous presidential power that Article 129 does not constrain when civilian authorities request assistance.

The governing jurisprudence long predated Calderón. Resolving acción de inconstitucionalidad 1/96 (March 1996, on the 1995 law that seated the SEDENA and SEMAR secretaries on the National Public Security Council), the SCJN held that the armed forces may constitutionally act in support of civilian authorities, at their express request, in public-security tasks, without a state of emergency or declaration of war, provided they remain subordinate to civilian authority and within legal bounds. The theses (P./J. 36/2000 through 38/2000 in their published form) became the legal floor under every joint operation of the Calderón and Peña Nieto eras [TBD-VERIFY: precise thesis numbers and publication dates]. Critics — among them a substantial constitutional-law literature and the CNDH — argued that the 1996 doctrine had been written for liaison and exceptional assistance and could not bear the weight of a permanent, nationwide military constabulary; the Court did not revisit the doctrine's core during the deployments' first decade. The result was the period's defining legal anomaly: the largest domestic military operation in post-revolutionary history proceeded for thirteen years (2006–2019) on a jurisprudential interpretation, with no statute regulating the use of force, detention, or command relationships — a vacuum the army itself repeatedly asked the political class to fill.

3.2 The 2017 Internal Security Law and Its Destruction

The first legislative answer was the Ley de Seguridad Interior, passed in December 2017 in the Peña Nieto sexenio's final year with army backing and over the objections of the CNDH, the UN human-rights system, and a broad civil-society front (the Seguridad Sin Guerra coalition). The law would have given the deployments a permanent statutory basis: "internal security" as a defined mission, presidential declarations activating military intervention, military command of operations, and information generated under the law classified as national-security material. Its opponents read it as the legalisation of a permanent military regime over public security and the inversion of the constitutional rule that the military supports rather than supplants civilian authority. In November 2018 — after the election, in the interregnum before AMLO took office — the SCJN struck the law down in its entirety in acción de inconstitucionalidad 6/2018 and its accumulated actions, holding that Congress lacked competence to legislate "internal security" as a public-security matter assigned to civilian institutions, and confirming the constitutional baseline: public security is civilian in character [TBD-VERIFY: ruling date 15 November 2018 and the vote]. The ruling closed the statutory route. What followed demonstrated that a constitutional supermajority could simply move the baseline.

3.3 The 2019 Guardia Nacional Amendment and Its Mutations

The 26 March 2019 constitutional reform — passed with near-unanimity, including opposition votes purchased by civilian-character guarantees — created the Guardia Nacional in Article 21: a public-security institution "of civilian character, discipline and command," attached to the civilian security ministry (SSPC). The civilian character was the opposition's price; the militarising content lived in the transitory articles. The fifth transitory authorised the president, for five years (to 26 March 2024), to "dispose of the permanent armed forces in public security tasks in an extraordinary, regulated, fiscalised, subordinated and complementary manner" — language imported nearly verbatim from the Inter-American Court of Human Rights' Alvarado Espinoza v. Mexico judgment (2018), which had set those adjectives as the conditions under which military participation in policing could be conventionally tolerable. Other transitories transferred the Federal Police's assets and personnel (the corporation was dissolved in 2019) and seconded military police and naval police elements to the new force.

The subsequent mutations form a five-step sequence. (i) The 11 May 2020 presidential accord invoked the fifth transitory to order the armed forces' participation in public security for the full window — turning a permissive exception into a standing instruction. (ii) The September 2022 ordinary-law reform transferred the GN's operational and administrative control to SEDENA — the de facto reality made de jure by statute. (iii) The SCJN struck that transfer down in April 2023 (acción de inconstitucionalidad 137/2022) as a fraud on the constitutional text's civilian-character command, ordering the GN's return to civilian control by January 2024 [TBD-VERIFY: ruling of 18 April 2023, 8–3 vote commonly reported]. (iv) Meanwhile the November 2022 constitutional amendment — passed with PRI support in the Plan-B political season — had extended the fifth transitory's armed-forces window from 2024 to 2028, with nominally strengthened reporting requirements [TBD-VERIFY: promulgation 18 November 2022]. (v) In September 2024, the incoming Morena supermajority elected in June — in the same constitutional-reform wave as the judicial reform (treated at MX-D-04) — amended Articles 13, 21, and related provisions to place the Guardia Nacional permanently under SEDENA as a military-discipline force with public-security functions, additionally empowering it to conduct investigations under prosecutorial authority [TBD-VERIFY: promulgation 30 September 2024, among AMLO's final acts in office]. The constitutional route thus consummated exactly what the Court had twice prohibited by inferior norms: the 2018 ruling was answered by the 2019 amendment's transitories, and the 2023 ruling by the 2024 amendment. The sequence is the period's cleanest illustration of a general proposition that recurs across the post-2024 corpus (MX-D-04, MX-I-01): against a supermajority that can amend the constitution at will, constitutional-review victories are provisional.

3.4 Military Jurisdiction: The Fuero Reform

The one sustained counter-militarising legal development of the period came from outside. Four Inter-American Court judgments against Mexico — Radilla Pacheco (2009, the dirty-war disappearance of Rosendo Radilla in Guerrero in 1974), Fernández Ortega and Rosendo Cantú (2010, the rapes of two indigenous Me'phaa women by soldiers in Guerrero in 2002), and Cabrera García and Montiel Flores (2010, the torture of two environmental activists) — held that the extension of military jurisdiction over human-rights violations against civilians violated the American Convention and ordered Article 57 of the Code of Military Justice reformed [TBD-VERIFY: case details and judgment dates]. The SCJN's landmark expediente Varios 912/2010 (July 2011), resolving how to comply with Radilla, bound all Mexican judges to conventionality control and to the restriction of the fuero; and in June 2014 Congress finally amended Article 57 to exclude crimes committed by soldiers against civilians from military jurisdiction, sending them to civilian courts. The reform was real but bounded: crimes by soldiers against soldiers, and the institution's internal opacity, remained untouched; civilian prosecutors and courts proved little more capable of convicting soldiers than military ones had been (Tlatlaya, §5.1, became the demonstration case); and the 2024 GN amendment, by re-housing a 100,000-strong police force within military discipline, re-raised the jurisdictional boundary questions the 2014 reform had been thought to settle — GN members' offences now sit ambiguously between the civilian regime the 2019 text promised and the military regime the 2024 text restored [TBD-VERIFY: implementing legislation's jurisdictional allocation for GN personnel].

4. The AMLO Expansion (2018–2024)

4.1 The Paradox Stated

No element of the AMLO record has generated more analytical commentary than the distance between the candidate and the president on the armed forces. Candidate López Obrador's position was the most demilitarising of any major contender in the post-2006 era: his campaign literature and speeches promised to return the army to its barracks within three years, condemned the Calderón deployment as a catastrophic error committed to legitimise a stolen election, and proposed amnesty-and-peace-process approaches to the conflict. President López Obrador, within months of taking office, declared that withdrawing the army was impossible given the condition of the police, built the Guardia Nacional on a military chassis, signed the May 2020 accord activating the maximum constitutional window for military policing, extended that window to 2028, transferred the GN to SEDENA by law and then — through his successor coalition — by constitution, and assigned the armed forces a portfolio of civilian functions without precedent in the institution's history. His own framing dissolved the contradiction rhetorically rather than substantively: the armed forces were pueblo uniformado, the people in uniform; their employment in development tasks was not militarisation but the recovery of a national asset; and the GN was the fulfilment, not the betrayal, of the barracks promise, since it was formally a new police rather than the army. The abrazos no balazos security doctrine and its operational record are treated at MX-J-02 §7; what this document tracks is the institutional construction that proceeded beneath the pacific rhetoric.

4.2 Building the Guardia Nacional

The GN was assembled at speed from three sources: the dissolved Federal Police (whose members joined under protest, with significant attrition), the Military Police, and the Naval Police — the latter two seconded en bloc while formally remaining military personnel, with military salaries, benefits, and discipline. Its first commander, Luis Rodríguez Bucio, was a retired army general; its operational culture, training pipeline, basing, and logistics were SEDENA's from the outset. By the end of the sexenio the force reported over 100,000 members deployed across some 260-plus regional coordinations [TBD-VERIFY: reported strength ranged from ~118,000 (2023) to ~130,000+ (2025); the share of members of military origin is commonly reported at 75–85%, with civilian-recruited members a minority]. Evaluations by México Evalúa, Causa en Común, and the WOLA-affiliated research community converged on three findings: the GN's certification, training, and complaint-handling lagged far behind its expansion; its detention-and-prosecution productivity per officer was low relative to the federal police it replaced [TBD-VERIFY: comparative metrics]; and its civilian character was a legal fiction years before the 2024 amendment formalised the fact. The counterfactual defence — that the Federal Police was itself corrupt (its last chief, Luna-era inheritance included), and that only the military institution could field a clean national force quickly — is the delivery argument of §4.5 applied to security.

4.3 The Mission-Creep Catalogue

The functional expansion beyond security is best apprehended as a catalogue; each line item carries its own decree, company, or accord.

  • Airports. Army engineers built the Felipe Ángeles International Airport (AIFA) at Santa Lucía — the replacement for the cancelled Texcoco airport (the cancellation itself treated in the MX-C-02 4T record) — delivering it in March 2022; a SEDENA-controlled company was created to operate it, and the military's airport-operating group subsequently accumulated a portfolio of additional airports including Tulum (built and opened 2023), Palenque, Chetumal, and others transferred from civilian operators [TBD-VERIFY: the full list under the Grupo Aeroportuario, Ferroviario, de Servicios Auxiliares y Conexos Olmeca-Maya-Mexica (GAFSACOMM), the SEDENA holding company consolidated in 2023].
  • The Tren Maya. SEDENA built several sections of the 1,500-kilometre tourist railway and emerged as its owner-operator through a SEDENA company (Tren Maya S.A. de C.V.), along with associated hotels built and run by the military [TBD-VERIFY: number of military-built hotels, commonly reported at six].
  • Customs and ports. A July 2020 decree transferred port authority and the merchant marine from the civilian transport ministry to SEMAR; in 2021 the customs system was transferred to military administration, with SEDENA and SEMAR officers running land-border and maritime customs respectively under the new national customs agency (ANAM) [TBD-VERIFY: ANAM creation and the division of customs posts].
  • An airline. The Mexicana de Aviación brand was revived as a SEDENA-owned commercial airline, launching operations in December 2023.
  • The welfare state's infrastructure. Army engineers built the Banco del Bienestar branch network — the physical infrastructure of the 4T's cash-transfer state (MX-G-02), with a target on the order of 2,700 branches — and military logistics ran the COVID-19 vaccine distribution (Operation Correcaminos) and portions of medicine distribution after the civilian procurement reforms faltered.
  • Pipelines, fuel, and the rest. Soldiers guarded Pemex pipelines in the 2019 anti-huachicoleo campaign (MX-K-02 carries the energy context); the armed forces ran the Islas Marías conversion, sargassum collection, tree-nursery logistics for Sembrando Vida, and stretches of migration enforcement — the 2019 GN deployment to the southern and northern borders under US tariff pressure (MX-E-01) being among the force's first mass employments.

The catalogue's analytical weight is cumulative: each item had a plausible ad-hoc justification, and the sum re-described the institution. By 2024 SEDENA was, by revenue and headcount, one of the largest infrastructure and transport conglomerates in the country.

4.4 Budget, Business, and Opacity

The expansion was fiscal as well as functional. SEDENA's budget roughly doubled in nominal terms across the sexenio, and the headline appropriations understate the flow: project transfers for AIFA and the Tren Maya, the GN's budget routed through military structures, and the creation of military-held trusts (fideicomisos) into which project revenues and appropriations were deposited placed substantial public money under military administration outside ordinary budget scrutiny [TBD-VERIFY: cumulative trust balances; investigative reporting (MCCI, Animal Político) and ASF audit observations documented multi-billion-peso trusts including the one funding military welfare from AIFA and Tren Maya revenues]. The November 2021 presidential accord (acuerdazo) declaring the government's priority projects matters of national security and public interest — read by INAI and civil society as an attempt to shield them from freedom-of-information and injunction litigation — was suspended in part by the SCJN, but national-security classification of military-run projects' contracts and studies persisted case by case [TBD-VERIFY: litigation outcomes; the SCJN invalidated the accord in 2023 in response to an INAI challenge, while project-level reservations continued]. The structural point exceeds any single reservation: the armed forces entered the 2020s with their historical opacity intact and their share of public resources transformed, producing an unprecedented quantum of unaudited state activity. The Superior Audit Office's observations on military-executed projects were repeatedly contested or ignored, and Congress — see §5.4 — never compelled resolution.

4.5 The Three Logics

Why did AMLO do it? The corpus holds three readings at strength, consistent with the three-account discipline.

The trust reading takes the president's words as the explanation. AMLO's expressed view — recurrent across the mañaneras (the anthology is a forthcoming Block-L entry) — was that the civilian state he inherited was structurally corrupt, that the neoliberal period had rotted the contractor class and the bureaucracy alike, and that the armed forces were pueblo uniformado: institutionally honest, nationalist, of humble social origin, and uncaptured by the oligarchy. Handing them the projects was, on this account, an integrity strategy. The reading's difficulty is the record: García Luna-era military protection scandals, Gutiérrez Rebollo, the Cienfuegos file, and the Guacamaya revelations all complicate the honesty premise, and the president's response to each was to defend the institution rather than test the premise.

The delivery reading is managerial and has substantial explanatory power. The army builds on schedule, does not strike, accepts direct presidential tasking without inter-ministerial negotiation, and operates under procurement and labour regimes far more permissive than civilian agencies'. For a president with a fixed six-year clock, signature projects that could not slip, and an expressed contempt for consultants and contractors, the military was the bureaucracy of last resort — the phrase from the comparative literature (§7.2) that Mexican analysts applied to the sexenio almost from its start. AIFA opened on time; Texcoco's civilian predecessor had not. The reading's difficulty is cost and quality: military execution was not audited, so the delivery claim was never priced.

The loyalty-purchase reading is political. A transformational president pursuing constitutional confrontation with the judiciary, the autonomous bodies, and the opposition had an interest in ensuring that the one institution capable of obstructing the project was invested in it instead. Missions, budgets, businesses, constitutional permanence, and conspicuous presidential deference — AMLO repeatedly shielded SEDENA from scandal, attacked its critics, and elevated its ceremonial standing — bound the armed forces to the 4T. The strong version (the army as veto player extracting rents from a dependent presidency) and the soft version (a prudent president coup-proofing a polarising transformation) bracket the same observation: by 2024 the armed forces had more to lose from the 4T's failure than from any conceivable alternative government's success. The reading's difficulty is the absence of evidence of demand-side extraction — the documented record shows presidential offer more than military demand, though the Guacamaya corpus complicates this [TBD-VERIFY: leaked-document evidence of SEDENA policy advocacy].

These logics are not exclusive, and the institutional outcome did not depend on which dominated: under any of them, the expansion ratcheted.

5. The Accountability Record

5.1 The Human-Rights Trajectory

The deployments' human-rights record is documented at MX-J-02 §5 in its conflict context; the institutional accounting is the concern here. CNDH complaints against SEDENA rose from low hundreds annually before 2007 to a peak in the upper 1,000s around 2009–2012, declining thereafter as the Navy's and then the GN's shares rose [TBD-VERIFY: CNDH series; SEMAR complaints spiked in the 2011–2012 marine-deployment period, and GN complaints climbed steadily from 2019]. The emblematic cases established the pattern of the accounting: in Tlatlaya (30 June 2014), soldiers killed 22 civilians in a warehouse confrontation in Mexico State, and the CNDH found that at least 12–15 were executed after surrender [TBD-VERIFY: CNDH Recommendation 51/2014 figures]; the soldiers prosecuted in civilian courts were acquitted or released, the military trial produced minimal sanction, and a documented military order to "kill criminals in the hours of darkness" (abatir delincuentes) was never traced up the chain of command [TBD-VERIFY: the order's text, revealed in Centro Prodh litigation]. The pattern — initial denial, CNDH recommendation, collapsed prosecution, no command responsibility — recurs across the period's catalogued cases (Tlatlaya; the 2010 Tamaulipas student killings; the 2017 Palmarito videotaped execution; the 2023 Nuevo Laredo killings by soldiers, which the AMLO government, unusually, sent to prosecution after CNDH intervention). The post-2014 jurisdictional reform (§3.4) changed the courtroom, not the outcome: a 2010s–2020s civilian-court conviction rate for soldiers in civilian-victim cases remained vanishingly low [TBD-VERIFY: WOLA and Centro Prodh prosecution-outcome tallies].

5.2 Ayotzinapa and the Archival Wall

The Ayotzinapa case (the full contested record at MX-J-01; the investigation's procedural history at MX-B-03) is this document's central accountability test because it placed a direct question — what did the army know and do on the night of 26–27 September 2014? — to an institution whose records could answer it. The documented elements: the 27th Infantry Battalion in Iguala monitored the students in real time through the C4 system and the army's own intelligence; a soldier had infiltrated the student body and was among the disappeared; and military personnel were present at scenes during the night. The AMLO-era truth commission (COVAJ, created 2018) and the special prosecutor (UEILCA, under Omar Gómez Trejo) pursued the military line that the original "historical truth" had been constructed to foreclose. The August 2022 COVAJ report named the events a "crime of state" with army participation; a general (José Rodríguez Pérez, the battalion's 2014 commander) and other military personnel were arrested. The unravelling followed within weeks: the FGR withdrew 21 of 83 arrest warrants including 16 against military personnel, Gómez Trejo resigned in September 2022 citing the destruction of his casework, and in July 2023 the GIEI international expert group ended its mandate with a final report stating that SEDENA and SEMAR had systematically withheld documents whose existence the experts had proven — including Iguala-night communications intercepts [TBD-VERIFY: precise warrant numbers, the August 2022 leaked-testimony controversies around the COVAJ report's evidentiary annex, and the GIEI's enumerated withheld documents]. AMLO's response to the collapse was to defend SEDENA publicly and assert that the army was cooperating fully. The episode's institutional meaning is stark: at the maximum point of presidential dependence on the armed forces, the presidency proved unwilling or unable to compel the institution's archives even for its own signature human-rights commitment — the dirty-war reflex of §2.3, operating intact in the 2020s.

5.3 Guacamaya: The Involuntary Audit

In September–October 2022 the hacktivist collective Guacamaya released roughly six terabytes of SEDENA internal email (2016–2022) — the largest documented breach of any Mexican state institution and, given the oversight vacuum, the closest thing to an external audit the institution has undergone. Reporting consortia (Latinus, Animal Político, Proceso, the international partners) extracted, among other findings: SEDENA's monitoring of journalists, human-rights defenders, and the Ayotzinapa families' lawyers, including the army's continued operation of Pegasus-type spyware after 2019 despite AMLO's pledge that state espionage had ended [TBD-VERIFY: the Ejército Espía investigation's attribution of post-2019 Pegasus infections to SEDENA]; military intelligence reporting on opposition and Morena politicians alike; internal documentation of the president's January 2022 health crisis beyond what had been disclosed; army knowledge of cartel-official collusion in specific states; and internal assessments candid about the GN's deficiencies [TBD-VERIFY: specific documents; the corpus treats Guacamaya-derived claims as journalistically mediated until the underlying documents are independently authenticated]. The government's response pattern matched §5.2: no prosecutions or dismissals arising from the revelations are documented, the president dismissed the surveillance findings, and SEDENA never publicly accounted for the breach itself — a security failure that, in a supervised institution, would itself have triggered external review.

5.4 The Oversight Vacuum

The civilian-oversight deficit is structural, not episodic. Congress's defence committees have no tradition of classified-session scrutiny, no professional staff capacity, and no practice of compelling military testimony; SEDENA secretaries appear for ceremonial comparecencias. The transparency regime's national-security exceptions — broad before 2018, broadened in practice by the megaproject classifications of §4.4 — exempted precisely the activities that expanded; and the 2024–2025 dissolution of INAI (the autonomous transparency body, absorbed into a government-controlled successor scheme in the same reform wave that consolidated the GN) removed the institution that had most consistently ordered military disclosures [TBD-VERIFY: INAI's extinction timeline and successor arrangements]. The ASF audits military-executed projects but cannot enforce; the CNDH under its post-2019 leadership largely ceased confronting the armed forces [TBD-VERIFY: the CNDH's recommendation rate against SEDENA/GN under Rosario Piedra Ibarra relative to predecessors]. The net structural fact: the institution whose functions, budget, and coercive presence in civilian life grew most across 2006–2026 is the institution over which every external-control mechanism weakened in the same period.

5.5 The Cienfuegos Affair

The arrest of General Salvador Cienfuegos Zepeda — SEDENA secretary 2012–2018 — at Los Angeles airport on 15 October 2020, on US federal charges (built through DEA interception of BlackBerry communications) of protecting the H-2 cartel, a Beltrán-Leyva successor organisation in Nayarit, under the alias "El Padrino," detonated the bilateral security relationship and exposed the bargain's deepest terms. The Mexican reaction was not merely diplomatic protest: Foreign Secretary Ebrard conveyed, by credible accounts, that DEA presence and cooperation in Mexico were at stake; on 17 November 2020 the US Attorney General moved to dismiss all charges and repatriated Cienfuegos in deference to "sensitive and important considerations" of the bilateral relationship; in January 2021 the FGR closed its own investigation within weeks, exonerating him, and AMLO ordered the DEA's evidentiary file published — an extraordinary act denounced by the US Justice Department — while declaring the US case fabricated; and in December 2020 Congress passed national-security-law amendments stripping foreign agents' immunities and restricting their operations [TBD-VERIFY: the sequence's specific dates; the published file's contents; the degree of subsequent DEA capability loss, which US officials later described as severe]. Three institutional readings coexist. For the army, the episode confirmed that the Mexican state — under the very president then maximising his dependence on it — would treat the institution's honour as a sovereign interest. For Washington, it demonstrated that no Mexican military figure could be pursued judicially without rupture, narrowing US options toward the unilateral instruments that surfaced in 2025 (§6.2, MX-E-02). For Mexican civil-military analysts, it answered the question the AMLO expansion had raised: asked to choose between accountability for the armed forces' apex and the armed forces' contentment, the civilian state chose contentment, unanimously, across both its diplomacy and its prosecution service. Whether Cienfuegos was guilty is precisely what no institution, in either country, finally established.

6. The Sheinbaum Continuity Question (2024–2026)

6.1 The Inheritance and the Dual Structure

Claudia Sheinbaum took office on 1 October 2024 the day after the GN-under-SEDENA constitutional amendment's promulgation — inheriting the militarised architecture not as a policy to continue but as a constitutional fact accomplished by her predecessor and ratified by her own coalition's votes. Nothing in her first two years reversed any element of it: the military enterprises operate and expand, the GN remains under SEDENA (now commanded through the defence secretariat under General Ricardo Trevilla Trejo's SEDENA), and the 2028 transitory horizon has been rendered moot by the permanence the 2024 amendment conferred [TBD-VERIFY: whether the 2028 armed-forces window retains independent legal significance post-amendment].

What Sheinbaum added is a civilian counter-model operating alongside military primacy rather than against it. Security Secretary Omar García Harfuch — the capital's former police chief, survivor of the 2020 CJNG assassination attempt, and the most empowered civilian security minister since the SSP's dissolution — was given what AMLO's civilian secretaries never had: operational leadership, presidential backing in public, and statutory tools. The 2024–2025 security-reform package restored investigation and intelligence powers to the civilian SSPC, created a national intelligence and investigation system under it, and rebuilt federal civilian investigative capacity dismantled in 2019 [TBD-VERIFY: the secondary legislation's content — the Ley del Sistema Nacional de Seguridad Pública reforms and the new intelligence-and-investigation law of 2024–2025]. The operational results — the raised tempo of arrests, fentanyl-lab seizures, and the unprecedented February and August 2025 transfers of cartel prisoners to the United States — are treated at MX-D-03; the institutional reading is what matters here: for the first time since 2006, the federal security strategy's public face and operational direction are civilian, while the deployed mass remains military. Whether this is the beginning of a rebalancing or a civilian veneer on a consolidated military structure is genuinely undetermined; the test cases — who controls the GN's tasking in practice, whether SSPC intelligence directs military operations or merely coexists with SEDENA's, and whether any military-held civilian function returns to civilian hands — have not yet resolved in either direction.

6.2 The External Compression: FTO Designations and the Sovereignty Red Line

The bilateral dimension transformed the institutional question in 2025. The second Trump administration designated six Mexican cartel organisations as Foreign Terrorist Organizations in February 2025 (MX-J-02 §8; MX-E-02 for the tariff linkage); administration figures publicly entertained unilateral US military action against cartel infrastructure; US surveillance flights near and over Mexican territory intensified; and reporting through 2025 described presidential direction to the Pentagon to prepare options against designated organisations, in a year when the US conducted lethal strikes against alleged trafficking vessels in the Caribbean and eastern Pacific [TBD-VERIFY: the scope of any 2025 directive concerning Mexican territory; the strike campaign through 2025–2026 was directed at maritime targets, with its extension toward Mexico debated but — as of this version date — not documented on Mexican soil]. Sheinbaum's response established the sexenio's sovereignty formula: "coordination yes, subordination no" (coordinación sí, subordinación no), explicit rejection of any US military intervention on Mexican territory, the February 2025 constitutional reform reinforcing sovereignty protections against foreign intervention [TBD-VERIFY: the reform's content, advanced after the FTO designations], and, in parallel, a deeper practical security cooperation than AMLO ever permitted — the prisoner transfers, expanded intelligence sharing, and a 2025 bilateral security framework re-institutionalising cooperation on mutually face-saving terms [TBD-VERIFY: the September 2025 framework's terms and nomenclature].

For the armed forces institutionally, the US pressure cuts both ways. It reinforces their indispensability — a sovereignty crisis is not the moment to shrink the army — and re-centres their external-defence identity for the first time in decades; SEDENA's planning discourse reportedly re-engaged territorial-defence scenarios long dormant [TBD-VERIFY]. But it also exposes the Cienfuegos legacy: US distrust of SEDENA specifically (the institutional memory of 2020, and the FTO framework's implicit treatment of Mexican state actors as potentially complicit) channels US cooperation toward Harfuch's civilian structures, which Washington visibly prefers as counterparts. The external pressure thus paradoxically strengthens the civilian counter-model inside the dual structure — an alignment of US preference and Mexican civilianist interest that no actor designed and either side may disavow.

6.3 Ratchet or Pendulum

The institutional question on which the period closes admits two scenarios, each with a coherent mechanism.

The restoration scenario — the pendulum — holds that Mexican civil-military history is a series of expansions and retractions around a stable subordination norm: the army entered public life massively in the Revolution and was returned to barracks by 1946; it entered the dirty war and receded; and the 2006–2024 expansion, driven by an emergency and an idiosyncratic presidency, can likewise recede as civilian police capacity is rebuilt (the Harfuch project's long logic), as the military enterprises are corporatised or divested under fiscal pressure, and as a future government — Morena or opposition — finds the military portfolio a liability. The scenario's evidence: the armed forces' own documented reluctance at various expansion stages (the army did not seek the drug war; its leadership repeatedly asked for legal cover rather than missions), the absence of any military political assertion across two decades of expansion, and the Sheinbaum administration's civilianising counter-currents.

The ratchet scenario — the new military state — holds that functional expansion is self-entrenching past a threshold Mexico has crossed. The mechanism is interests: a military that operates airports, railways, customs, an airline, and a 100,000-member police force has revenue streams, career structures, and budget claims that any retraction must expropriate; a political class that has dismantled the civilian alternatives (the Federal Police dissolved, INAI extinguished, the judiciary restructured) has nowhere to return functions to; and a constitution amended to make the arrangement permanent requires a supermajority to unwind — a supermajority the arrangement's beneficiary coalition controls. On this reading the 1946 comparison fails because the post-revolutionary demilitarisation traded the generals out of politics with a growing economy's spoils, while the current arrangement trades the institution into the economy itself. The scenario's evidence: no function transferred to the military between 2006 and 2026 has ever been transferred back [TBD-VERIFY: candidate counter-examples — partial civilian re-staffing of customs administration was reported in 2023–2024 after corruption episodes].

The corpus does not adjudicate. It registers the asymmetry of the stakes: if the pendulum reading is right, Mexico's anomalous twentieth-century achievement — a subordinated military in a region of praetorians — survives augmented by a constabulary episode; if the ratchet reading is right, the achievement has been quietly traded away by the civilians it protected, without a coup, without a crisis, by accumulation of decrees.

7. Comparative Dimension

7.1 The Latin American Militarisation Wave

Mexico's expansion is the largest instance of a regional pattern. Brazil's Garantia da Lei e da Ordem (GLO) deployments put the army into Rio's favelas and national strikes repeatedly from the 1990s through the 2018 federal intervention in Rio de Janeiro; the Bolsonaro presidency (2019–2022) appointed thousands of military personnel to civilian posts, including a general as health minister mid-pandemic, and the 8 January 2023 episode tested — and on the conventional reading confirmed, narrowly — the high command's electoral subordination. El Salvador under Bukele constitutes the region's pole: the army doubled in size, soldiers became the instrument of the territorial-control plan and the post-March-2022 régimen de excepción, and the February 2020 military entry into the Legislative Assembly displayed coercion in direct service of executive politics — the threshold Mexico has not approached. Guatemala, Honduras, and Ecuador (the 2024 "internal armed conflict" declaration) all re-militarised policing in the same generation; Colombia is the instructive contrast — a military larger and more combat-engaged than Mexico's throughout the period, but embedded in a constitutional and bilateral (Plan Colombia-era) accountability architecture, a strong constitutional court, and a peace-process framework that kept its mission bounded and its civilian control under continuous institutional negotiation, the falsos positivos scandal notwithstanding (the Colombia corpus treats this arc in its own block). The comparative literature (Diamint, Pion-Berlin, the WOLA and Latinobarómetro monitoring series) reads the wave's common drivers as the triad of criminal-violence emergencies, weak and distrusted police, and publics that rank the armed forces among their most trusted institutions — conditions under which assigning soldiers to civilian tasks is, for any incumbent, the politically cheapest available act, every time, regardless of cumulative cost.

7.2 The Theoretical Frame

The civil-military-relations literature offers three lenses, each partially fitting. Huntington's classic distinction (The Soldier and the State, 1957) between objective control — professionalising the military within a bounded military sphere — and subjective control — binding it to the rulers — renders the Mexican twentieth century legible as subjective control par excellence (loyalty to the PRI presidency, not to a civilian-supervised sphere), and the AMLO era as its renewal under a new hegemon: the 4T did not build objective control instruments; it re-bound the institution to the project. Alfred Stepan's framework (Rethinking Military Politics, 1988) — military role expansion and the distinction between the military as institution and as government, developed for bureaucratic-authoritarian Brazil — supplies the sharper warning: Stepan's "new professionalism of internal security" describes precisely the doctrinal drift by which constabulary missions expand a military's conception of its own competence until national development itself appears as a military task. The 2006–2026 mission set — from counter-narcotics to airports to railways to banks — is role expansion in textbook sequence, accomplished without the praetorian politics Stepan's cases exhibited. The third lens, associated with David Pion-Berlin and the contemporary Latin Americanists, is the military as bureaucracy of last resort: in states with weak civilian implementation capacity, elected governments rationally assign the military whatever must actually get done — disaster relief, vaccines, infrastructure — and such employment, the argument runs, need not threaten democratic control if missions are bounded and temporary. Mexico is this literature's limit case: the thesis explains every individual assignment and is silenced by their sum, since nothing in the period was bounded and nothing proved temporary.

7.3 What Distinguishes Mexico: Militarisation Without Militarism

The Mexican case's distinguishing configuration is the coexistence of three facts. First, no praetorian tradition: no successful coup since the institutionalisation of the 1920s, no military caudillo in politics since mid-century, no documented episode — through four alternations of party in the presidency (2000, 2012, 2018, 2024) — of the armed forces contesting an electoral outcome or conditioning a transfer of power. Second, electoral subordination intact: the army executed the orders of PAN, PRI, and Morena presidents successively, including orders each incoming president issued against its predecessor's policies. Third, functional expansion unprecedented: by 2026 the armed forces' share of state functions — security, borders, customs, ports, airports, railways, an airline, banking infrastructure, welfare logistics — exceeds that of any Latin American military not actually governing, and does so with the region's weakest external-oversight architecture (§5.4).

The formulation that has organised the Mexican debate — militarisation without militarism — names this configuration: the state and public life are militarised in function while the military remains non-militarist in politics. Its proponents (and the governments of the period, implicitly) draw the reassuring conclusion: the subordination norm is the real safeguard, it has held for eighty years, and functional employment does not erode it. The critics' rejoinder, grounded in the §5 record, is threefold. First, the formulation mistakes the absence of coups for the presence of control: the Mexican military does not need to seize power it is being handed. Second, militarism is not only praetorianism — an institution that surveils journalists, withholds evidence from prosecutors, and is shielded by the state from foreign and domestic justice is already exercising political power, in the form of impunity, whatever its electoral behaviour. Third, the configuration is metastable: subordination held under a hegemonic civilian force (the PRI, then the 4T) that could credibly reward loyalty; it has never been tested by a weak, divided, or hostile civilian government attempting retraction — the test a future alternation would administer, with the institution's accumulated interests then in play. The Mexican case thus closes where the comparative literature is thinnest: there is no developed theory of how a non-praetorian military with a praetorian-scale functional portfolio behaves when asked to give it back.

8. Conclusion

The arc admits a compact statement. A revolutionary army was traded out of politics by 1946 on terms — autonomy and opacity for subordination — that held for sixty years because the civilians never needed the soldiers for anything the bargain did not cover. In December 2006 a civilian president needed them for a war, and the bargain began repricing: missions for legal cover, deployments for budgets, indispensability for deference. Between 2019 and 2024 a second civilian president, elected to end the war, instead generalised the bargain from security to the state itself — and constitutionalised the result. At no point did the army take anything; at every point it was given, by civilians acting within their lawful powers, for reasons (an emergency, a distrusted bureaucracy, a transformation to insure) that were individually defensible and cumulatively transformative. The 2026 institution — a constitutionally entrenched security force of over 100,000 under military command, a construction-and-transport conglomerate, the least-audited large organisation in the Mexican state, and still, by every survey, among the most trusted — is the bargain's compound interest.

The forward questions for subsequent versions of this document: the practical command relationship between Harfuch's civilian intelligence structure and SEDENA's deployed mass, and any documented instance of a military-held function returning to civilian hands; the GN's jurisdictional and disciplinary regime as the 2024 amendment's implementing legislation is litigated; the fate of the military enterprises — GAFSACOMM, Tren Maya S.A., Mexicana — under fiscal pressure, and the first genuine audit of any of them; the Ayotzinapa military prosecutions and whether any command-level accountability ever materialises; the US dimension — whether the strike discourse crosses onto Mexican territory, and what that event would do to both the sovereignty consensus and the armed forces' internal politics; and the deepest line, untestable until tested: what the institution does the first time a Mexican government asks for something back. The twentieth century's pact was judged a success because that question never had to be asked. The twenty-first century's settlement has ensured that, sooner or later, it will be.

Sources

  1. Constitución Política de los Estados Unidos Mexicanos — Articles 13, 21, 89-VI, and 129; reform decrees of 26 March 2019 (Guardia Nacional), 18 November 2022 (armed-forces window extension), and September 2024 (GN under SEDENA) (Diario Oficial de la Federación).
  2. Suprema Corte de Justicia de la Nación: acción de inconstitucionalidad 1/96 (1996) and derived theses; acción de inconstitucionalidad 6/2018 and accumulated (November 2018, Ley de Seguridad Interior); acción de inconstitucionalidad 137/2022 (April 2023, GN–SEDENA transfer); expediente Varios 912/2010 (2011, Radilla compliance).
  3. Inter-American Court of Human Rights: Radilla Pacheco v. Mexico (2009); Fernández Ortega et al. v. Mexico and Rosendo Cantú et al. v. Mexico (2010); Cabrera García and Montiel Flores v. Mexico (2010); Alvarado Espinoza et al. v. Mexico (2018).
  4. Ley de la Guardia Nacional (2019) and reform decrees; Ley de Seguridad Interior (2017, invalidated); Acuerdo presidencial of 11 May 2020 (armed forces in public security); Acuerdo of 22 November 2021 (priority projects as national security).
  5. Camp, Roderic Ai, Generals in the Palacio: The Military in Modern Mexico (Oxford University Press, 1992), and Mexico's Military on the Democratic Stage (Praeger, 2005).
  6. Serrano, Mónica, "The Armed Branch of the State: Civil-Military Relations in Mexico," Journal of Latin American Studies 27:2 (1995).
  7. Benítez Manaut, Raúl, multiple works on Mexican national security and civil-military relations, including the CASEDE Atlas de la Seguridad y la Defensa de México series (2009–2024).
  8. Moloeznik, Marcos Pablo, and María Eugenia Suárez de Garay, on the militarisation of public security in Jalisco and nationally, Frontera Norte and related venues.
  9. Pérez Ricart, Carlos A., Cien años de espías y drogas (Debate, 2022), on the drug-mission history of Mexican security institutions.
  10. Comisión Nacional de los Derechos Humanos (CNDH), annual complaint statistics against SEDENA, SEMAR, and Guardia Nacional, 2006–2025; Recommendation 51/2014 (Tlatlaya).
  11. Comisión para el Acceso a la Verdad, el Esclarecimiento Histórico y el Impulso a la Justicia de las violaciones graves a los derechos humanos 1965–1990, reports and public statements (2021–2024); FEMOSPP draft historical report (2006, as leaked).
  12. Comisión para la Verdad y Acceso a la Justicia del Caso Ayotzinapa (COVAJ), report of August 2022; GIEI (Grupo Interdisciplinario de Expertos Independientes), Ayotzinapa reports I–VI (2015–2023), especially the final report's findings on SEDENA/SEMAR document withholding.
  13. WOLA (Washington Office on Latin America), reports on militarised policing, the Guardia Nacional, and military justice in Mexico (2010–2025); Centro Prodh, Tlatlaya: A un año and related litigation documentation.
  14. México Evalúa and Causa en Común, evaluations of the Guardia Nacional's composition, certification, and performance (2019–2025).
  15. Mexicanos Contra la Corrupción y la Impunidad (MCCI), Animal Político, and Quinto Elemento Lab investigations into military trusts, AIFA/Tren Maya contracting, and SEDENA business structures (2020–2025).
  16. Reporting consortium analyses of the Guacamaya leaks (Latinus, Proceso, Animal Político, R3D's Ejército Espía on Pegasus attribution), 2022–2023.
  17. US Department of Justice, United States v. Salvador Cienfuegos Zepeda (E.D.N.Y., 2020), indictment and dismissal filings; FGR, exoneration determination and published case file (January 2021).
  18. Storrs, K. Larry, and subsequent Congressional Research Service reports on US–Mexico security cooperation; the Mérida Initiative and Bicentennial Framework documentation (2008–2025).
  19. Huntington, Samuel P., The Soldier and the State (1957); Stepan, Alfred, Rethinking Military Politics: Brazil and the Southern Cone (1988); Pion-Berlin, David, and Rafael Martínez, Soldiers, Politicians, and Civilians: Reforming Civil-Military Relations in New Democracies (2017); Diamint, Rut, "A New Militarism in Latin America," Journal of Democracy 26:4 (2015).
  20. Guevara Moyano, Íñigo, and International Institute for Strategic Studies (IISS) Military Balance entries on Mexican force structure and budgets, 2006–2026.
  21. Press and analytical record 2006–2026: Reforma, El Universal, Milenio, Proceso, Nexos, Letras Libres, El País México, and Animal Político coverage of the deployments, the GN's construction, the military-enterprise expansion, the Ayotzinapa military prosecutions, the Cienfuegos affair, and the 2025–2026 US pressure sequence.
  22. Sheinbaum administration security-reform legislation and mañanera record (2024–2026); SSPC and SEDENA public reporting under the Harfuch strategy.
  • MX-I-01: The INE and Mexican Electoral Institutional Architecture (1990–2026) — the parallel autonomous-institution arc; the supermajority-versus-institutional-design problem common to both documents
  • MX-J-02: The Mexican Drug War — Cartel Evolution, Militarisation, and the Security Debate (2006–2026) — the conflict narrative this document's institutional analysis presupposes; the militarisation paradox at its §7
  • MX-A-02: Calderón Presidency (2006–2012) — the December 2006 deployment decision in its presidential context
  • MX-B-03: Ayotzinapa — the Procedural Investigation (2014–2024) — the investigation arc whose military dimension is treated at §5.2
  • MX-J-01: Ayotzinapa 43 — the Contested Record — the full case treatment, including the army's documented night-of presence
  • MX-C-02: AMLO and the Fourth Transformation (2018–2024) — the 4T programme of which the military expansion was the implementing arm
  • MX-D-03: Sheinbaum Security Policy and the Sinaloa Cartel War (2024–2025) — the Harfuch strategy and operational record inherited at §6.1
  • MX-E-02: US–Mexico Trump-2 Tariff Confrontation and the 2025–2026 Trade Shock — the FTO-designation and sovereignty-pressure context of §6.2
  • MX-K-02: The 2013 Energy Reform and Its Reversal (2013–2026) — the huachicoleo pipeline-guarding mission and the Pemex context of §4.3
  • MX-N-01: Mexico in International Perceptions — the external reading of the militarisation within the cartel imaginary
  • MX-O-01: Mexico Megatrends — The 2030s Questions
  • MX-G-03: Mexican Water Governance — Scarcity, Concessions, and the Crisis of the 2020s
  • MX-M-01: "The Perfect Dictatorship" — PRI Hegemony as Political Idea
  • MX-F-02: Mexico-China Relations — Competitor, Supplier, and the Nearshoring Triangle
ArchiveSourcesChat