CO-G-02: The Coca Economy and Drug Policy — From Fumigation to Total Peace (1994–2026)

Status: DRAFTWords: 8,149

1. Key Takeaways

  • Coca cultivation in Colombia is a rational household response to state absence, not primarily a criminal choice. The crop concentrates in the roadless agrarian periphery — Nariño, the Catatumbo region of Norte de Santander, Putumayo, and Cauca, which together account for roughly 70 per cent of national cultivation [TBD-VERIFY: precise departmental shares in the most recent UNODC Monitoreo] — where coca dominates legal alternatives on every margin that matters to a smallholder: the buyer travels to the farm (eliminating the transport cost that makes cacao or coffee uneconomic on unpaved mule tracks), the crop yields four to six harvests per year, pasta base functions as a non-perishable store of value and de facto currency, and armed-group buyers extend planting credit that no bank will. Thirty years of policy that attacked the plant without altering these margins is the structural explanation for the cultivation record.

  • The UNODC cultivation series describes an inverse-U followed by an unprecedented second peak. Measured cultivation stood at approximately 163,000 hectares in 2000, fell to a trough of approximately 48,000 hectares in 2012–2013 — the headline claim of the spraying-era "success" — then rebounded continuously: 146,000 (2016), 171,000 (2017), 204,000 (2021), 230,000 (2022), and a record of approximately 253,000 hectares in 2023, the highest figure ever recorded for any country [TBD-VERIFY: the 2024 measurement year figure, published by UNODC-SIMCI in late 2025, and whether it confirms stabilisation or further growth]. Potential cocaine output rose even faster than area — approximately 2,664 metric tons in 2023 against roughly 690 tons in 2000 [TBD-VERIFY: 2000-era potential-production baseline] — because yields per hectare and laboratory extraction efficiency improved across the period.

  • The fumigation era (1994–2015) sprayed approximately 1.8 million cumulative hectares and is the most studied eradication programme in drug-policy history; the academic consensus is that it failed on cost-effectiveness grounds. Aerial glyphosate spraying, authorised by the Consejo Nacional de Estupefacientes in 1994 and industrialised under Plan Colombia from 2000 (peak: ~172,000 hectares sprayed in 2006), displaced cultivation rather than eliminating it — the intra-Colombian "balloon effect" that moved coca from Putumayo and Caquetá to Nariño, the Pacific littoral, and Catatumbo. Econometric work (Mejía, Restrepo, Rozo) estimated that one hectare of spraying reduced net cultivation by a small fraction of a hectare at a cost per kilogram of cocaine suppressed that was an order of magnitude above interdiction alternatives [TBD-VERIFY: precise elasticity estimates — commonly cited range 0.02–0.065 ha reduction per ha sprayed in the Mejía-Restrepo work].

  • The 2015 suspension of aerial spraying was a health-and-law decision, not a drug-policy decision, and the distinction conditioned everything after. The March 2015 IARC reclassification of glyphosate as "probably carcinogenic to humans" (Group 2A) gave the Santos government the scientific predicate; the Consejo Nacional de Estupefacientes suspended spraying in May 2015 (operations ceased by October). The Corte Constitucional then constitutionalised the suspension: Sentencia T-236 de 2017 and Auto 387 de 2019 imposed precautionary-principle conditions — independent health evidence, prior consultation with affected communities, a regulatory body independent of the spraying agencies — that no subsequent government has satisfied. The Duque administration's 2019–2021 restart attempt (including Decreto 380 de 2021) died on these conditions, which means Colombia's most politically demanded eradication instrument has been judicially unavailable for over a decade.

  • The 2016 peace agreement's Point 4 created the largest voluntary-substitution experiment ever attempted — and its underfunding is the central policy failure of the post-accord decade. The Programa Nacional Integral de Sustitución de Cultivos de Uso Ilícito (PNIS), created by Decreto 896 de 2017, enrolled approximately 99,000 families who committed to eradicate in exchange for a staged two-year payment package (approximately COP 36 million per family) plus productive-project support. UNODC verification found compliance above 90 per cent — enrolled families genuinely eradicated [TBD-VERIFY: precise verified-compliance and replanting figures, commonly cited as ~94–98 per cent eradication compliance and under 1 per cent replanting among participants]. But the state defaulted on its half of the bargain: by the early 2020s only a small minority of families had received productive projects, the programme stopped enrolling new families in 2018, and the fiscal cost of full implementation was never appropriated. The lesson — substitution works at the household level and fails at the fiscal-political level — structures the entire current debate.

  • Coca territories became the killing fields of the post-accord period. The FARC's 2017 demobilisation removed the monopoly taxing-and-governing authority from the coca economy without replacing it with the state; ELN, the Estado Mayor Central (EMC) dissidence, the Segunda Marquetalia, and the Clan del Golfo (AGC) fought to inherit the gramaje taxation, purchase networks, and corridors. Social-leader and ex-combatant assassinations concentrated overwhelmingly in coca municipalities — PNIS local leaders were systematically targeted precisely because substitution threatened armed-group revenue [TBD-VERIFY: Indepaz and UN Verification Mission cumulative figures for social-leader killings 2016–2025 and the share in coca municipalities]. The coca economy is therefore not only a drug-policy problem but the financial substrate of Colombia's post-FARC conflict ecology (CO-E-01, CO-D-03).

  • Petro's 2023–2033 drug policy is the most explicit paradigm rejection by any producer-country government in the prohibition era. Framed by his 26 September 2022 UN General Assembly address declaring the war on drugs "irrational" and a failure, the policy — Sembrando vida, desterramos el narcotráfico (October 2023) — formalised two axes: "oxygen" for cultivating territories (voluntary, gradual substitution; no forced eradication of smallholder plots; coca-leaf industrialisation pilots) and "asphyxia" for trafficking structures (interdiction, laboratory destruction, asset seizure, targeting of finance and corridors rather than farmers). Manual-eradication targets were cut from the Duque-era 130,000 hectares to roughly 20,000; cocaine seizures rose to record levels (approximately 740+ tons in 2023 and higher again in 2024 [TBD-VERIFY: official seizure statistics for 2023–2025]). The policy's defenders read record seizures plus reduced state-farmer violence as vindication; its critics read the 2023 cultivation record as refutation.

  • The 2023 cultivation record admits two honest readings, and the corpus records both. The policy-failure reading: removing eradication pressure while substitution remained underfunded predictably allowed expansion, and the Petro government enrolled almost no new families into substitution while Paz Total ceasefires gave armed groups freedom to expand the economy they tax. The measurement-and-lag reading: the 2023 figure reflects planting decisions taken in 2021–2022 under Duque; cultivation had risen under every post-2014 government regardless of paradigm; UNODC's own reporting shows cultivation concentrating (more density in fewer enclaves) rather than spreading, and prices for coca leaf collapsed in 2022–2024 from oversupply — meaning the binding constraint was never Colombian eradication policy but global demand and trafficking-network purchasing. Both readings are empirically grounded; neither is decisive on present evidence.

  • The trafficking structure has fragmented and de-nationalised even as supply boomed. The vertically integrated Medellín and Cali cartels gave way after the mid-1990s to successive fragmentations — Norte del Valle, paramilitary trafficking, post-2006 bacrim — leaving the AGC/Clan del Golfo as the dominant Colombian trafficking organisation but with Mexican organisations (Sinaloa, CJNG) controlling wholesale purchase and Balkan/Albanian networks the booming European market. Export routes shifted toward the Pacific (semi-submersibles, container contamination through Ecuadorian ports), making Ecuador's post-2021 security collapse substantially a Colombian-supply spillover. The September 2025 US decertification of Colombia (the determination of 15 September 2025, with a national-interest waiver — CO-F-03 §10) and the Trump-2 administration's Cartel de los Soles allegations against Venezuela embedded the coca question in the hemisphere's most adversarial geopolitics since the 1990s.

  • At the 2026 election, four policy positions contend, and thirty years of evidence under-determines the choice among them. The re-securitisation position (the uribista candidates) promises restored aerial spraying and eradication targets — against the judicial conditions of T-236/2017 and the cost-effectiveness literature. The continuity position defends Petro's substitution-and-interdiction framework — against the 2023 cultivation record and the fiscal non-delivery of substitution. The regulation position (the Global Commission lineage, the COP16-era Cali debates, the coca-leaf industrialisation pilots) argues prohibition itself is the policy failure — against the reality that no consumer state will legalise cocaine. The rural-development-first technocratic consensus argues that only roads, land titling, and state presence change cultivation decisions — against a 15-year horizon and a fiscal cost no government has paid. What the evidence supports is narrower than what any position claims: forced eradication without development does not durably reduce cultivation; substitution works where funded and protected; interdiction raises trafficker costs without reducing area; and cultivation responds most strongly to prices, security vacuums, and state presence — variables drug policy only partially controls.

2. The Coca Political Economy

2.1 Where coca grows and why

Coca cultivation in Colombia is not randomly distributed; it maps with near-perfect fidelity onto the geography of state absence. The four departments that have dominated cultivation since the mid-2010s — Nariño (the Pacific littoral and the Andean piedmont around Tumaco), Norte de Santander (the Catatumbo region on the Venezuelan border), Putumayo (the Amazonian piedmont bordering Ecuador), and Cauca (the Pacific slope and the Micay canyon) — share a profile: colonisation frontiers settled in the twentieth century by displaced and land-poor migrants, minimal road connectivity, weak or absent land titling, chronic armed-group presence, and distances to legal markets that render most licit agriculture uneconomic. Tumaco municipality alone has at various points contained more coca than the whole of Bolivia [TBD-VERIFY: comparative figure by year — Tumaco peaked near 19,000–23,000 ha in 2017–2018 against Bolivian national totals of ~23,000–29,000 ha]. UNODC's Monitoreo de Territorios Afectados por Cultivos Ilícitos has repeatedly found that roughly half of all cultivation sits within "special management zones" — resguardos indígenas, Afro-Colombian collective-title territories (consejos comunitarios), forest reserves, and national-park buffers — where both eradication and development intervention face legal and consultative constraints [TBD-VERIFY: precise share by report year].

The household economics are the irreducible core of the problem, and they have been documented consistently across three decades of fieldwork. For a smallholder on an unpaved mule track six hours from the nearest market town, coca dominates every legal alternative on the margins that matter. The buyer comes to the farm gate — the raspachín harvest labour market and the pasta base purchase network eliminate the transport cost that makes cacao, coffee, or plantain uneconomic at distance. The crop yields four to six harvests a year against one or two for most licit perennials, and begins producing within twelve to eighteen months of planting. Pasta base — the intermediate coca paste produced on-farm or in nearby cocinas — is compact, non-perishable, and functions in many coca municipalities as a de facto currency and store of value in economies where banks do not exist. Armed-group and trafficker buyers extend planting credit and guarantee purchase at announced prices; no agricultural bank or licit crop buyer does either. The standard finding of household surveys is that coca farmers are not wealthy — gross household incomes from coca typically sit near or modestly above the rural poverty line [TBD-VERIFY: UNODC/DNP household income survey figures] — but coca is the only crop that reliably monetises at all. The policy implication, repeated in every serious evaluation since the 1990s, is that cultivation decisions respond to the relative profitability and risk of coca against alternatives, and that the state can alter that calculus either by raising the cost of coca (eradication, which thirty years of evidence shows works weakly and temporarily) or by raising the return to legality (roads, titles, markets, security — which no government has funded at scale).

2.2 The cultivation series, 1994–2026

Two measurement systems track Colombian coca — the UNODC-SIMCI satellite census (conducted with the Colombian government) and the US government's ONDCP estimate — and they diverge systematically, with the US series typically 20–40 per cent higher. The corpus uses the UNODC series as the Colombian-official baseline and flags divergences.

The UNODC arc is an inverse-U followed by an unprecedented second mountain. Cultivation climbed through the late 1990s as Peruvian and Bolivian suppression (the "air bridge denial" era) relocated the industry to Colombia — the original, hemispheric balloon effect — reaching approximately 163,000 hectares in 2000 [TBD-VERIFY: UNODC 2000 baseline; figures of 160,000–163,000 ha are commonly cited]. The Plan Colombia spraying decade drove the measured series down to a trough of approximately 48,000 hectares in 2012–2013, the figure on which the entire "supply-control success" narrative rests. From 2014 the series rebounded without interruption under three governments of opposed paradigms: ~96,000 (2015), ~146,000 (2016), ~171,000 (2017), a brief plateau under Duque's eradication surge, then ~204,000 (2021), ~230,000 (2022), and approximately 253,000 hectares in 2023 — the highest cultivation figure UNODC has ever recorded for any country [TBD-VERIFY: all series figures against UNODC-SIMCI annual reports; the 2024 measurement, published late 2025, and whether it shows stabilisation]. Potential cocaine production rose faster still — to approximately 2,664 metric tons in 2023 [TBD-VERIFY] — because the industry's productivity improved: higher-yielding varietals, denser planting, more harvests per year, and better laboratory extraction meant each hectare produced roughly twice the cocaine of a 2000-era hectare.

The drivers of the post-2014 rebound are multiple and partially confounded, which is why the series resists single-cause readings. The 2015 spraying suspension removed the principal eradication instrument. The Havana negotiations created an anticipation effect — communities planted coca expecting substitution payments to be allocated by hectarage [TBD-VERIFY: extent documented in UNODC and FIP reporting]. The 2014–2016 collapse in gold and oil prices pushed informal labour back toward coca. The peso's depreciation raised the local-currency value of a dollar-denominated crop. And the FARC's withdrawal from territorial governance after 2016 opened a competitive scramble among successor groups whose start-up financing was coca. Disentangling these is an active econometric literature; the honest summary is that cultivation rose under spraying's absence, but had already begun rising before the suspension and rose fastest in zones the spraying era had never controlled.

2.3 Armed-group taxation across eras

The coca economy has been taxed and governed by every armed actor in the conflict, and the modalities matter for policy. The FARC-EP institutionalised the gramaje — a per-kilogram levy on pasta base, on laboratory operation, and on shipments transiting its territory — from the 1980s, alongside price-setting, dispute adjudication among growers and buyers, and prohibitions on growers consuming the product. By the 2000s coca revenue was the FARC's largest income stream [TBD-VERIFY: estimates ranged from US$200 million to over US$1 billion annually depending on method and motive of the estimator]. The AUC paramilitaries were more directly integrated into trafficking — several bloc commanders were traffickers who had purchased franchises — and their 2003–2006 demobilisation seeded the bacrim successor organisations. The ELN, historically ambivalent about drug money on doctrinal grounds, abandoned the ambivalence in the 2010s: its expansion in Catatumbo and along the Venezuelan border, and its January 2025 offensive against FARC dissidents in Catatumbo (CO-E-01), are widely read as wars for coca rent. The post-2016 landscape — EMC dissidence under Iván Mordisco, the Segunda Marquetalia under Iván Márquez, the Clan del Golfo/AGC, and the ELN — reproduces FARC-style taxation in fragmented, competitive form. The governance consequence is central to this document's argument: where one armed group holds a monopoly, the coca economy is orderly and violence is low; where succession is contested, coca municipalities become the most violent in the country. Drug policy that alters the economy without anticipating the armed-group response — as both the 2016 accord and Paz Total arguably did — changes the conflict as much as the crop.

3. The Fumigation Era (1994–2015)

3.1 The architecture of aerial eradication

Aerial spraying of illicit crops predates the coca era — marijuana plantations in the Sierra Nevada were sprayed with paraquat and then glyphosate from 1978 — but the systematic aerial campaign against coca began with the Consejo Nacional de Estupefacientes (CNE) authorisation of 1994, under the Gaviria-Samper transition and under heavy US pressure during the narcocasete crisis of Samper's legitimacy. The programme (Programa de Erradicación de Cultivos Ilícitos con Glifosato, PECIG) was operated by the Anti-Narcotics Directorate of the National Police with US-contracted aircraft, pilots, and chemical supply — DynCorp's spray fleet flying from bases at Tumaco, San José del Guaviare, and Larandia became one of the most visible artefacts of US presence in the country (CO-A-01).

Plan Colombia industrialised the programme from 2000. Annual sprayed area rose from roughly 58,000 hectares in 2000 to a peak of approximately 172,000 hectares in 2006 [TBD-VERIFY: official PECIG spraying statistics by year], with cumulative spraying over the programme's life commonly estimated at 1.6–1.8 million hectares — an area larger than some Colombian departments, much of it sprayed repeatedly. The spray mixture was glyphosate at concentrations and application rates above normal agricultural use, with the surfactant Cosmo-Flux added to improve leaf adhesion [TBD-VERIFY: formulation details from CNE technical documents]. Spraying was conducted at fixed-wing speed and altitude over smallholder landscapes where coca interplanted with food crops, which guaranteed — whatever the toxicology — that licit crops, pasture, and water sources were repeatedly hit. Compensation procedures for wrongful spraying existed on paper and almost never paid out [TBD-VERIFY: approval rates for compensation claims, commonly reported below 5 per cent].

3.2 Did it work? The effectiveness debate

The fumigation programme is the most studied supply-control intervention in the history of drug policy, and the econometric verdict is unusually consistent. The headline correlation favoured the programme: national cultivation fell from ~163,000 hectares (2000) to ~48,000 (2012–13) across the high-spraying decade. But the causal decomposition did not. The work of Daniel Mejía, Pascual Restrepo, and Sandra Rozo — using spray-exclusion zones around the Ecuadorian border and other identification strategies — estimated that one hectare sprayed reduced net cultivation by only a small fraction of a hectare [TBD-VERIFY: commonly cited point estimates of 0.02–0.065 ha per ha sprayed], because growers replanted, pruned and resuscitated sprayed bushes, moved plots under forest canopy, or shifted municipalities. The implied cost per kilogram of cocaine removed from the market via spraying ran to multiples — by some estimates an order of magnitude — of the cost via interdiction or via demand-side treatment in consumer countries [TBD-VERIFY: Mejía-Restrepo cost-effectiveness estimates, RAND demand-side comparisons].

The balloon effect operated at every scale. Hemispherically, Colombia's boom was itself the displacement of Peruvian and Bolivian suppression. Intra-nationally, the spraying decade moved coca out of Putumayo and Caquetá — the early Plan Colombia "push into southern Colombia" theatres — into Nariño, the Pacific littoral, Catatumbo, and across the borders into Ecuador and Venezuela's border fringe. The departments that dominate cultivation in the 2020s are substantially the departments the spraying era created as coca frontiers. Defenders of the programme make a narrower claim that survives the econometrics: spraying raised costs and risk enough to contain cultivation below its unconstrained ceiling, and the post-2015 rebound is the counterfactual made visible. That claim is not refutable on present evidence, but neither does it rescue the cost-effectiveness case, and it must be weighed against what containment purchased in legitimacy terms — which is the subject of the next subsection.

3.3 Health, legality, and the contestation

From its earliest years the programme generated complaints — skin and respiratory ailments, miscarriages, destroyed food crops, poisoned fish ponds — that the toxicological literature could neither cleanly confirm nor dismiss, since glyphosate's agricultural-use safety profile said little about repeated aerial application over inhabited smallholdings. Ecuador sued Colombia at the International Court of Justice in 2008 over border-zone spraying; Colombia settled in 2013, paying US$15 million and accepting a 10-kilometre no-spray strip [TBD-VERIFY: settlement terms]. Domestically, the Corte Constitucional had been circling the programme for a decade — Sentencia SU-383 de 2003 ordered prior consultation with indigenous communities of the Amazon before spraying their territories.

The end came through health science rather than drug policy. In March 2015 the WHO's International Agency for Research on Cancer reclassified glyphosate as Group 2A — "probably carcinogenic to humans." The Santos government, then deep in the Havana negotiations and already committed in draft Point 4 to a substitution-first paradigm, took the IARC finding as sufficient predicate: the Minister of Health recommended suspension, and the CNE voted in May 2015 to suspend aerial glyphosate spraying, with final operations ceasing by October 2015. Critics noted the convenience of the timing — the FARC had long demanded an end to spraying — and the scientific contestation that followed IARC's finding (regulatory agencies including EPA and EFSA did not concur) kept the toxicology politically alive.

What converted a suspension into a durable prohibition was the Court. Sentencia T-236 de 2017 held that any restart required satisfaction of the precautionary principle: objective and conclusive evidence of safety evaluated by an authority independent of the spraying agencies, prior consultation with affected ethnic communities, and a complaint-and-compensation mechanism that actually functioned. Auto 387 de 2019 elaborated the conditions. The Duque government's restart attempt — culminating in Decreto 380 de 2021 re-establishing the regulatory framework — proceeded through environmental licensing at ANLA but never satisfied the consultation requirements before the government's term expired; Petro's government formally abandoned the restart in 2022. The cumulative effect is that Colombia's most politically demanded eradication instrument has been judicially unavailable since 2015, and any 2026 government promising its return must either satisfy T-236's conditions — which a decade of effort has not — or confront the Court (CO-I-01).

4. The Peace Agreement Experiment (2016–2022)

4.1 Point 4 of the Havana accord

The 2016 Acuerdo Final devoted its fourth chapter — "Solución al Problema de las Drogas Ilícitas," agreed in principle at Havana in May 2014 — to the most explicit paradigm statement any Colombian government had signed (CO-B-02). Point 4 framed coca cultivation as a social phenomenon rooted in rural poverty and state absence, to be addressed primarily through voluntary substitution integrated with the comprehensive rural reform of Point 1; it committed the state to differentiate between growers (subjects of development policy) and traffickers (subjects of criminal policy); it promised a public-health approach to consumption; and it subordinated forced eradication to cases where communities refused substitution agreements. The FARC, for its part, committed to end all involvement in the drug economy and to contribute to substitution — a commitment whose breach by dissident factions became one of the implementation era's defining facts.

4.2 PNIS: design and the implementation record

The Programa Nacional Integral de Sustitución de Cultivos de Uso Ilícito (PNIS), created by Decreto 896 de 2017, operationalised Point 4. Families signing individual agreements committed to eradicate their coca within an agreed schedule in exchange for a staged package totalling approximately COP 36 million per family over two years [TBD-VERIFY: package composition — commonly described as ~COP 12 million in immediate food-security payments across the first year, plus self-sustaining project support (~COP 1.8 million), short-cycle project support (~COP 9 million), and a productive project (~COP 10 million), with technical assistance]. Enrolment reached approximately 99,000 families across 56 municipalities in 14 departments before the Duque government closed new enrolment in 2018 [TBD-VERIFY: precise enrolment figure — 98,000–99,097 families is the commonly cited range], with collective framework agreements signed by communities representing far more.

The implementation record splits cleanly into a success and a failure, and both are essential to the 2026 debate. The success: UNODC verification found compliance among enrolled families above 90 per cent — commonly reported at 94–98 per cent verified eradication, with replanting below 1 per cent among participants [TBD-VERIFY: UNODC PNIS verification reports], a result without precedent in forced-eradication zones, where replanting historically ran above a third. Substitution, in other words, worked at the household level when the household believed the bargain. The failure: the state defaulted on its side. Payments arrived late and incompletely; technical assistance reached a fraction of families; and by the early 2020s only a small minority of enrolled families had received the productive projects that were the package's entire developmental logic [TBD-VERIFY: Contraloría and Kroc Institute figures — productive-project delivery was commonly reported in single-digit or low-double-digit percentages as late as 2021–2022]. The fiscal cost of honouring all signed agreements — variously estimated above COP 4 trillion — was never fully appropriated by any government. The PNIS thus became the policy world's most cited natural experiment: proof that voluntary substitution can achieve near-total compliance, and proof that the Colombian state has not yet paid for it.

4.3 The killing of the substitution leadership

The post-accord security collapse in coca territories (CO-D-03, CO-E-01) fell with particular precision on the people PNIS created. Local substitution committees required visible community leaders to organise enrolment, verify eradication, and press the state for delivery — and those leaders threatened, by function, the revenue of every armed group taxing the local coca economy. Social-leader assassinations after 2016 concentrated overwhelmingly in coca and illegal-mining municipalities, with PNIS leaders and ex-FARC combatants among the most targeted categories [TBD-VERIFY: Indepaz cumulative counts — over 1,400 social leaders killed 2016–2024 on Indepaz figures, with UN Verification Mission counts lower on stricter methodology; the share in PNIS municipalities]. The murders operated as policy: they raised the personal cost of joining substitution above any payment package, and in several regions enrolment collapsed after leader killings. Any account of why substitution "failed to scale" that omits the assassination campaign is incomplete — the programme was underfunded and its local cadre was being killed.

4.4 The Duque interregnum

The Duque government (2018–2022; CO-C-01) inherited a programme its coalition had campaigned against and a cultivation curve already at record levels. Its response was a paradigm reversal within the accord's letter: PNIS enrolment was frozen (existing agreements honoured slowly, no new families), forced manual eradication was surged to record levels — approximately 130,000 hectares manually eradicated in 2020, the highest ever [TBD-VERIFY] — and the aerial-spraying restart was pursued through Decreto 380 de 2021. The results sharpened every prior lesson: manual eradication at record scale coincided with cultivation remaining at or near record levels, replanting in forcibly eradicated plots ran far above PNIS replanting rates, and eradication operations in communities awaiting unpaid PNIS packages produced repeated confrontations, blockades, and deaths of both civilians and eradicators [TBD-VERIFY: casualty figures in eradication confrontations 2019–2021]. By the end of the Duque term, the policy stalemate was complete — spraying judicially blocked, substitution fiscally defaulted, forced eradication demonstrably non-durable — and it was into this stalemate that Petro's paradigm arrived.

5. Petro's Paradigm Shift (2022–2026)

5.1 The declaration: "the war on drugs has failed"

On 26 September 2022, six weeks into his presidency, Gustavo Petro used his first UN General Assembly address to deliver the most explicit repudiation of the prohibition paradigm any sitting producer-country president had made: the war on drugs had "failed," had cost a million Latin American lives [TBD-VERIFY: verbatim phrasing of the figure in the UNGA text], and had treated the Amazonian peasant as an enemy while the consumption that drove the market sat untouched in the global North. He pressed the theme — addiction as a disease of loneliness and consumer-society pathology, prohibition as a machine that converts US demand into Colombian graves — in language closer to manifesto than diplomacy (CO-D-02). The speech mattered less as policy than as signal: Colombia, prohibition's most decorated soldier, was defecting from the paradigm's rhetoric, and every subsequent bilateral friction with Washington (CO-F-03) traces partly to that declaration.

5.2 The 2023–2033 policy: oxygen and asphyxia

The formal policy followed in October 2023: the National Drug Policy 2023–2033, Sembrando vida, desterramos el narcotráfico ("Sowing life, we banish narco-trafficking"). Its architecture rested on two announced axes. "Oxygen" for the cultivating territories: no forced eradication of smallholder plots as the default instrument; gradual, negotiated, collective substitution tied to territorial transformation; legal pathways for coca leaf in industrial, food, and traditional uses (fertiliser, beverages, the cannabis-adjacent regulatory debates); and an explicit distinction between the grower household and the trafficking economy. "Asphyxia" for the trafficking structure: interdiction of cocaine and precursor chemicals, destruction of cristalizaderos (crystallising laboratories), asset forfeiture, targeting of finance, corridors, and armed-group command layers. Forced manual eradication targets were cut from the Duque-era ~130,000 hectares to roughly 20,000 hectares annually [TBD-VERIFY: announced versus achieved eradication 2023–2025], concentrated on industrial-scale and armed-group-owned plantations rather than smallholdings.

The interdiction half of the bargain produced its headline numbers: cocaine seizures rose to records — approximately 740+ metric tons in 2023, with 2024 reported higher again [TBD-VERIFY: Ministry of Defence seizure statistics 2023–2025; the 2025 figure amid decertification politics] — alongside record laboratory destructions. The substitution half largely did not materialise as programme: the government inherited the unpaid PNIS roll, prioritised paying down its arrears [TBD-VERIFY: extent of PNIS arrears cleared under Petro], and launched pilot territorial-transformation programmes (Catatumbo, Micay) without a successor enrolment at PNIS scale. Critics across the spectrum converged on this point: whatever the paradigm's merits, the Petro government did not fund or build the substitution apparatus its own policy made central.

5.3 The 2023 record and the two readings

The UNODC's October 2024 report measuring 2023 — approximately 253,000 hectares, potential production ~2,664 tons — landed as the empirical crisis of the paradigm, and the corpus records both honest readings (Key Takeaway 8). The failure reading is straightforward: the government removed eradication pressure, enrolled almost no new substitution families, and presided over Paz Total ceasefires (CO-D-03, CO-D-04) that gave the EMC, ELN, and AGC freedom to expand the economy they tax; cultivation duly hit the highest level ever recorded. The lag-and-structure reading is equally grounded: coca takes one to two years from planting decision to satellite-measurable maturity, so the 2023 measurement substantially reflects 2021–2022 decisions under Duque's surge; cultivation rose under every government and every paradigm after 2014; UNODC's own analysis showed cultivation concentrating into fewer, denser enclaves (the share of cultivation in Nariño, Cauca, Putumayo, and Norte de Santander rising) rather than metastasising nationally; and the 2022–2024 coca-leaf price collapse — farm-gate pasta base prices falling by half or more in several regions, with unsold harvests reported in Putumayo and Guaviare [TBD-VERIFY: price-series data from UNODC and field reporting] — indicated that the binding constraint on the economy was trafficking-network purchasing capacity and global demand, not Colombian eradication. On the second reading, Colombia's cultivation statistics measure the world cocaine market's appetite more than they measure Colombian policy. Both readings will be tested by the 2024 and 2025 measurement years; neither was decisively confirmed by mid-2026 [TBD-VERIFY: 2024 UNODC figure].

5.4 Total Peace and the trafficking groups

Paz Total extended the negotiation instrument from political insurgencies to what the law had previously treated as pure crime: alongside the ELN and EMC tables, the government opened "socio-legal conversations" with the Clan del Golfo and urban trafficking structures (Medellín, Buenaventura). For drug policy the doctrine cut both ways. Its theory held that negotiated submission or transformation of trafficking groups could demobilise the coca economy's armed superstructure wholesale — the only intervention scale that matches the problem. Its record by 2026 was largely adverse: ceasefires reduced state-group combat while groups expanded territorial control and coca taxation; the EMC table fractured and partially returned to war (the Cauca and Catatumbo crises of 2024–2025, CO-D-04, CO-E-01); and no trafficking structure reached a submission agreement with judicial architecture to receive it [TBD-VERIFY: status of the AGC sujeción framework and the Buenaventura/Medellín processes as of mid-2026]. The drug-policy verdict on Paz Total therefore mirrors the verdict on the 2016 accord: the diagnosis (the coca economy is governed by armed groups, so drug policy is conflict policy) was sound; the execution did not hold the territory the diagnosis identified.

5.5 The American rupture

The Petro paradigm collided with the most drug-hawkish US administration in decades. Across 2024–2025 the bilateral relationship (CO-F-03) deteriorated through migration and tariff disputes into the drug question itself: on 15 September 2025 the Trump administration decertified Colombia as a counter-narcotics partner — the first decertification since 1997 — citing record cultivation, while issuing a national-interest waiver that preserved most assistance flows [TBD-VERIFY: against the CO-D-08/CO-F-03 corpus line on the decertification determination, waiver scope, and any subsequent change through mid-2026]. The decertification's practical bite was narrower than its symbolism, but the symbolism was the point: Washington formally declared the Petro paradigm a failure, Petro answered in kind, and US strikes on alleged trafficking vessels in the Caribbean and the Cartel de los Soles designation against the Maduro government (CO-F-02) wrapped Colombian drug policy into the hemisphere's sharpest confrontation since Plan Colombia's founding era — this time with Bogotá and Washington on opposite sides of the argument. For the 2026 candidates, the relationship with Washington and the drug paradigm are now a single question.

6. Interdiction, Cartels, and Trafficking Structure

6.1 From cartels to fragmentation

The trafficking industry that sits atop the coca economy has passed through three structural generations, each produced by the state's victory over the last. The Medellín and Cali cartels (roughly 1976–1995) were vertically integrated national enterprises controlling the chain from purchase through US wholesale; their destruction — Escobar's death in December 1993, the Cali leadership's capture in 1995 — was the founding success of US-Colombian cooperation and the founding demonstration of its limits, since cocaine flows did not fall. The second generation — the Norte del Valle cartel, the trafficking wings of the AUC paramilitary blocs, and the FARC's deepening role — was smaller, more violent internally, and more entangled with the armed conflict. The third generation emerged from the 2003–2006 paramilitary demobilisation: the bandas criminales ("bacrim"), of which the Clan del Golfo (self-styled Autodefensas Gaitanistas de Colombia, AGC; also "Urabeños") became the consolidated survivor — by the 2020s the largest armed trafficking organisation in the country, present in hundreds of municipalities, running extortion, illegal mining, and migrant smuggling (the Darién) alongside cocaine [TBD-VERIFY: AGC strength estimates — commonly 6,000–9,000 armed members plus support networks in security-force reporting]. The 2021 capture and 2022 extradition of its commander Otoniel (Dairo Antonio Úsuga) — answered by a four-day paro armado that paralysed much of northern Colombia — demonstrated both the state's reach and the organisation's resilience: the AGC replaced him without visible disruption.

The strategic consequence of three decades of "kingpin" success is an industry that no longer has Colombian kingpins. Mexican organisations — Sinaloa and CJNG — moved from buyers to principals, controlling wholesale purchase in Colombia and the US market's economics; Balkan and Albanian networks built direct purchase relationships for the European market, often stationing buyers in Colombian port cities; Colombian groups increasingly function as armed suppliers and logistics contractors in a chain whose profits concentrate abroad. Fragmentation also dispersed the violence: the post-cartel industry kills fewer judges and ministers and more rural social leaders.

6.2 The supply boom and the routes

The 2014–2024 decade produced the largest cocaine supply expansion in history — cultivation roughly quintupling from the 2013 trough while yield gains compounded the area growth. Global demand absorbed it: European consumption surged (Antwerp and Rotterdam seizure records year after year), new markets opened in Latin America, West Africa, Asia, and Oceania, and purity rose while retail prices held or fell across consumer markets [TBD-VERIFY: EMCDDA/UNODC World Drug Report market indicators] — the standard market evidence that interdiction, however successful in tonnage, was not binding on supply.

Routes shifted with enforcement. The Caribbean "go-fast" era gave way to the Pacific: semi-submersibles and low-profile vessels built in the mangrove yards of Nariño and Cauca carrying multi-ton loads toward Central America and Mexico, and increasingly on long-haul runs toward Oceania and even trans-Atlantic routes [TBD-VERIFY: documented long-range semi-submersible interdictions]. Container contamination through legal trade became the volume channel to Europe — through Cartagena, Buenaventura, and above all through Ecuador: Guayaquil's banana trade, dollarisation, weak port controls, and the post-2016 vacuum on the Colombian side of the border made Ecuador the principal export platform for Nariño-Putumayo cocaine. Ecuador's security collapse after 2021 — prison massacres, the assassination of presidential candidate Fernando Villavicencio in 2023, the January 2024 declaration of "internal armed conflict" — is substantially a downstream effect of the Colombian supply boom meeting Mexican and Balkan purchasing networks on Ecuadorian soil. Venezuela's role — transit corridor for Catatumbo product, refuge for ELN and Segunda Marquetalia command, and the contested Cartel de los Soles allegations against the Maduro state — embedded the routes question in the 2025–2026 US-Venezuela confrontation (CO-F-02).

6.3 The interdiction record

Interdiction is the one instrument whose output rose under every government: seizures climbed from roughly 100–200 tons annually in the early 2000s to ~500 tons by 2021 and approximately 740+ tons in 2023, with 2024–2025 reported at or above that level [TBD-VERIFY: seizure series; Colombian seizures plus international seizures of Colombian-origin product were estimated by some analysts to exceed half of potential production, a share without precedent]. The honest reading of the record is double-edged. Seizures at this scale impose real costs — they are the empirical core of Petro's "asphyxia" claim and of the price-collapse evidence in §5.3, and some analysts attribute the 2022–2024 farm-gate price fall partly to interdiction pressure on purchasing networks. But thirty years of seizure growth coincided with thirty years of supply growth; traffickers price interdiction losses as a cost of business; and the literature's standard finding is that interdiction raises trafficker costs and violence along routes without durably reducing area planted or consumer-market availability. Interdiction, like eradication, is a holding instrument, not a solution — the difference is that it holds without spraying villages.

6.4 Precursors, money, and the upstream-downstream asymmetry

Two further interdiction fronts complete the structural picture, and both illustrate the same asymmetry. Precursor chemicals — gasoline, cement, and sulphuric acid for pasta base; potassium permanganate and solvents for crystallisation — flow into coca regions through ordinary commerce at volumes that controlled-substance regimes have never meaningfully constrained; laboratory destruction (thousands of cocinas and hundreds of cristalizaderos annually [TBD-VERIFY: laboratory-destruction statistics by year]) imposes replaceable losses on an industry whose fixed capital is trivial relative to margins. Money is the asymmetry's sharpest expression: the overwhelming share of the cocaine value chain's profit is realised in consumer markets and laundered through their financial systems — the standard estimates leave producer-country actors with a single-digit share of final retail value [TBD-VERIFY: value-chain distribution estimates] — yet enforcement effort and enforcement deaths concentrate almost entirely at the Colombian end. Colombian asset-forfeiture and anti-laundering capacity has grown across the period (the SAE, the UIAF, the extinción de dominio regime), but the financial architecture that converts Colombian coca into European and North American wealth sits outside Colombian jurisdiction. Every Colombian government since Samper has made some version of this point to its US counterpart; the 2022–2026 difference is that Petro made it the organising frame of national policy rather than a diplomatic aside.

7. The Policy Debate at 2026

Four positions contend in the 2026 election cycle (CO-D-07, CO-D-09, CO-D-10), each with genuine evidence behind it and a genuine evidentiary problem in front of it. The corpus presents them analytically; the record under-determines the choice.

The re-securitisation position — carried by the uribista and right-of-centre candidacies — promises restored aerial spraying, forced-eradication targets in six figures, military confrontation with the EMC, ELN, and AGC, and repaired alignment with Washington. Its evidence: the 2012–13 trough remains the only sustained cultivation decline in the series, and it occurred under maximum-pressure policy; the Petro-era pairing of reduced eradication and expanded armed-group control is the most recent natural experiment, and cultivation hit its record within it. Its problems: the trough's causal attribution to spraying is precisely what the econometrics undermined; the instrument it promises is judicially unavailable absent satisfaction of T-236's conditions or a confrontation with the Constitutional Court; and the social cost of the eradication-confrontation cycle in communities owed unpaid substitution packages is a documented driver of rural alienation in exactly the municipalities the state needs to win.

The continuity position defends the Petro framework — substitution-first, interdiction against structures, no war on growers — arguing it needs funding and time, not reversal. Its evidence: PNIS compliance rates remain the strongest single result in the field's history; record seizures and the farm-gate price collapse are consistent with effective "asphyxia"; cultivation concentration rather than spread suggests the frontier is no longer expanding. Its problems: the paradigm's own substitution apparatus was never built at scale under the government that proclaimed it; the 2023 record is, at minimum, not evidence of success; and the position's viability depends on a bilateral relationship with Washington that decertification has placed at its lowest point since Samper.

The regulation position — the Global Commission on Drug Policy lineage, voiced in Colombia by figures across the Santos and Petro worlds and aired at the COP16 Cali debates and in coca-leaf industrialisation pilots — holds that prohibition itself is the policy failure: the price premium that makes coca dominant at the farm gate, finances every armed group, and corrupts every institution is a creature of illegality, and no eradication or interdiction mix alters it. Its evidence is the entire thirty-year record this document describes. Its problem is jurisdictional: cocaine regulation requires consumer-state and treaty-system movement that does not exist and shows no sign of existing; unilateral Colombian steps are confined to coca-leaf derivatives and symbolic treaty diplomacy [TBD-VERIFY: status of the Colombian-Bolivian push on coca-leaf review at the UN treaty bodies]. Regulationists answer that the same was said of cannabis a decade before Canada and half the United States; the analogy's force is contested.

The rural-development-first position — the technocratic consensus of the evaluation literature, the Kroc Institute reporting, and much of the planning establishment — holds that cultivation responds durably only to the variables that change the household calculus: roads, land titling, market access, schools, and physical security; that Point 1 of the 2016 accord (comprehensive rural reform), not Point 4, was always the real drug policy; and that every eradication paradigm fails because it attacks the symptom of state absence. Its evidence is the consistency of the household economics across every era (§2.1) and the PNIS demonstration that credible state bargains produce compliance. Its problems are the horizon and the price: territorial transformation operates on 15-year timescales no electoral cycle rewards, at fiscal costs — the accord's rural-reform chapter alone was costed in the tens of trillions of pesos — that no government, of any orientation, has yet appropriated, in territories where armed groups kill the programme's local cadres.

What thirty years of evidence actually supports is narrower than any position's platform: forced eradication without development does not durably reduce cultivation; voluntary substitution achieves compliance where funded and protected, and has never been funded and protected at scale; interdiction raises trafficker costs without reducing area or availability; aerial spraying is the least cost-effective instrument and is judicially foreclosed; and cultivation responds most strongly to prices, security vacuums, and state presence — variables that drug policy, narrowly conceived, only partially controls. The 2026 election will choose a rhetoric; whether it chooses a different result depends on whether any government pays for the rural state that every paradigm, at bottom, presupposes.

8. Conclusion

The coca economy is the longest-running policy problem in modern Colombian governance, and its thirty-year record has a structure worth stating plainly. Every paradigm — aerial spraying (1994–2015), forced manual eradication (continuous, surged 2018–2022), voluntary substitution (2017–), interdiction-first "asphyxia" (2022–) — has been tried at scale; none has durably reduced cultivation; and the series' only sustained decline (to the 2012–13 trough) remains causally contested between the spraying that claimed it and the conflict, price, and displacement dynamics that confound it. Meanwhile the variables that demonstrably move cultivation — farm-gate prices, armed-group governance, the credibility of state bargains, the existence of roads — sit mostly outside the drug-policy toolkit and inside the larger questions of rural state-building that the 2016 accord diagnosed and no government has funded.

The document's second conclusion concerns honesty under uncertainty. The 2023 cultivation record genuinely admits two readings — paradigm failure and measurement lag atop a demand-driven market — and the 2024–2025 measurement years, the post-decertification bilateral trajectory, and the fate of the substitution arrears will adjudicate between them only partially. The corpus holds both readings open, as it holds open the four-cornered 2026 debate, because the evidence does. What the evidence does not hold open is the proposition, asserted in every era's launch rhetoric, that the next instrument will succeed where its predecessors failed without changing the household economics of the cultivating periphery. Thirty years in, that proposition has been tested to destruction. The coca question at 2026 is therefore not which eradication instrument Colombia chooses next; it is whether any Colombian government — of left or right, aligned with or estranged from Washington — will pay for the rural state whose absence the coca bush has measured, hectare by hectare, since 1994.

Sources

  1. UNODC-SIMCI, Monitoreo de Territorios Afectados por Cultivos Ilícitos (annual series, 2001–2024; especially the October 2024 report measuring 2023).
  2. Gobierno de Colombia / FARC-EP, Acuerdo Final para la Terminación del Conflicto y la Construcción de una Paz Estable y Duradera, Chapter 4 (24 November 2016).
  3. Ministerio de Justicia y del Derecho, Política Nacional de Drogas 2023–2033: Sembrando vida, desterramos el narcotráfico (October 2023).
  4. Decreto 896 de 2017 (creating PNIS) and Decreto 380 de 2021 (aerial-spraying regulatory framework).
  5. Corte Constitucional, Sentencia SU-383 de 2003, Sentencia T-236 de 2017, and Auto 387 de 2019.
  6. IARC, Monograph 112: Glyphosate (March 2015 Group 2A classification).
  7. Mejía, Daniel, and Pascual Restrepo, "The Economics of the War on Illegal Drug Production and Trafficking," Journal of Economic Behavior & Organization (2016), and related working papers on spraying effectiveness.
  8. Rozo, Sandra V., "Is Murder Bad for Business? Evidence from Colombia," and aerial-spraying evaluation work (World Bank / UCLA working papers).
  9. Mejía, Daniel, "Plan Colombia: An Analysis of Effectiveness and Costs," Brookings Institution (2016).
  10. UNODC, Informes de verificación PNIS (PNIS verification report series, 2018–2022).
  11. Kroc Institute for International Peace Studies, Comprehensive Accord Implementation Reports (annual, 2017–2024), Point 4 chapters.
  12. Fundación Ideas para la Paz (FIP), reports on coca, eradication confrontations, and Paz Total security dynamics (2017–2025).
  13. Indepaz, registries of social-leader and ex-combatant assassinations (2016–2025).
  14. International Crisis Group, Deeply Rooted: Coca Eradication and Violence in Colombia (Report No. 87, 2021) and subsequent Colombia briefings (2022–2025).
  15. WOLA (Washington Office on Latin America), Colombia drug-policy monitoring and decertification analyses (Isacson et al., 2015–2025).
  16. Gootenberg, Paul, Andean Cocaine: The Making of a Global Drug (University of North Carolina Press, 2008).
  17. Tate, Winifred, Drugs, Thugs, and Diplomats: U.S. Policymaking in Colombia (Stanford University Press, 2015).
  18. Petro, Gustavo, Address to the 77th UN General Assembly (26 September 2022).
  19. The White House / US Department of State, Presidential Determination on Major Drug Transit or Major Illicit Drug Producing Countries (15 September 2025) and prior annual determinations.
  20. UNODC, World Drug Report (annual, 2015–2025), cocaine market chapters.
  21. ICJ, Aerial Herbicide Spraying (Ecuador v. Colombia) — application (2008) and settlement (2013).
  22. Contraloría General de la República, fiscal audits of PNIS implementation (2019–2023).
  • CO-A-01 — Plan Colombia: the US-assistance and counter-insurgency framework within which the spraying era was financed and flown.
  • CO-B-02 — The Havana negotiations: where Point 4's substitution paradigm was agreed.
  • CO-B-03 — The 2016 plebiscite and implementation era: the political environment that conditioned PNIS funding.
  • CO-C-01 — The Duque presidency: the eradication surge and the spraying-restart attempt.
  • CO-D-03 — Total Peace doctrine and the post-2024 trajectory: the negotiation framework governing the armed groups that tax coca.
  • CO-D-04 — Petro's second half: the Paz Total stalemate, the Catatumbo and Cauca crises.
  • CO-E-01 — The ELN cycle: the Catatumbo war for coca rent.
  • CO-F-02 — Colombia–Venezuela: trafficking corridors, ELN refuge, and the Cartel de los Soles confrontation.
  • CO-F-03 — Colombia–US relations: certification politics, Plan Colombia diplomacy, and the 2025 decertification.
  • CO-G-01 — The Colombian economic model: the macroeconomy within which the coca economy and rural underdevelopment sit.
  • CO-G-03: back-reference added by symmetry sweep
  • CO-K-02: back-reference added by symmetry sweep
  • CO-N-01: back-reference added by symmetry sweep
  • CO-O-01: back-reference added by symmetry sweep
ArchiveSourcesChat