The Failing War on Drugs Network Model of Drug Flow from Mexico to the U.S. LCDR Ben Cipperley CPT Kevin Larrabee 11 JUN 2012
Feb 21, 2016
The Failing War on Drugs
Network Model of Drug Flow from Mexico to the U.S.
LCDR Ben CipperleyCPT Kevin Larrabee
11 JUN 2012
Agenda
• Problem Statement• Drug War Background• Modeling• Producer/ Consumer Model• Demand Flow Model• Conclusions• Questions
Problem Statement• 65% of narcotics in the U.S. flow from Mexico
– $45 billion annually in producer/distributor profits– 99% of all flow from Mexico is overland (U.S. DOJ)
• Understand the overland flow of drugs network from Mexico into the U.S.– How do narcotics flow across the border?– What motivates flow?
• How to best attack the network– Attack Demand– Attack Distribution– Attack Supply
War on Drugs
• U.S. War on Drugs– Roots formed during Prohibition (1920’s)– DEA created (1973)– CIA and military involvement (1982)
• Demand topped $60 billion in 2012
• $15 billion on drug enforcement in 2010– ~$500 per second– Interdiction budget $3.9 billion
Front Lines
• U.S. Customs and Border Patrol– 21,444 Agents– 18,506 on the SW border (86.3%)
• Apprehensions– 96% of all CBP apprehensions
occur on the southwest US Border
• Military Planning Model– Info and Intel driven response
A Typical Border Crossing
• Different capacities by location– San Ysidro, CA = 24 lanes – Laredo, TX = 20 lanes– Naco, AZ = 2 lanes
• 2.7 million loaded container trucks entered the U.S. through the SW border in 2009
• An estimated 30% of all narcotics shipments are seized at the border
De La Vega Background
• Don Diego de la Vega is the new drug lord of the Zetas Cartel and is focused on pushing cocaine into the U.S.
• Zetas cartel struck new deal to combine resources with Sinaloa Federation and the Cartel Pacifico Sur
• Cocaine Manufacturing facilities emerge in 8 Mexican cities
Real Life ScenarioThe Drug Lord- Don Diego De La Vega
Objectives
• Understand how to best attack the network by measuring the amount of profit gained
• Red Team (De La Vega) – Maximize profit through minimum cost flow– Maximize narcotics flow into the U.S.
• Blue Team (U.S. Border Patrol)– Maximize the cost of production/distribution– Reduce the amount of narcotics flow– Improve interdiction capability
Model
Model
Model with Network Overlay
Ciudad Victoria
END
START
Modeling the Network
• Assumptions– Omniscient and omnipotent Drug Lord– Shipments occur overland using legal U.S. border
crossings• Both the U.S. and de la Vega act optimally• Flow starts from Mexico (no domestic production)• Simultaneous attacks• Limitations– Only modeled legal border crossings– No political effects considered– Blue Force interdiction cost not calculated
Model Evolution
• Shortest Path– How does cocaine flow across the border?
• Producer/Consumer– Basic supply/demand model– Interdiction lowers capacity to zero– Not accurate for how drugs move (NOT Wal-mart!)
• Demand Flow– Profit motivates flow– Every crossing evaluated for “efficiency”– Interdiction lowers efficiency to zero– Demand vs. Supply Interdiction– Reduction in efficiency through training
Producer/ Consumer
Producer/Consumer
• Supply of 9,500 kilos produced across nine facilities in Mexico
• Demand of 9,500 kilos in the U.S.• Costs to use a road are based upon price of fuel
per gallon and distance to destination• Costs to move flow through a border crossing are
bribe amounts as a function of number of arrests• Interdictions reduce the flow capacity to zero by
making the cost equal infinity
Producer/ Consumer ModelSUPPLY
Ciudad Victoria -500Saltillo -1000Torreon -500Monterrey -2000Reynosa -600Matamoros -900Chihuahua -2500Hermosillo -1500
DEMANDBakersfield 215San Diego 794Los Angeles 3289Phoenix 1075Tucson 252Austin 440Dallas 1634Houston 1525San Antonio 549
• Supply is a function of projected production facilities
• Demand is a function of population percentage of the Southwest
Additional ParametersPRODUCTION COSTS
Ciudad Victoria 12000Saltillo 15500Torreon 12500Monterrey 15000Reynosa 13000Matamoros 13500Chihuahua 12000Hermosillo 11000
BRIBESSanYsidro 1295 Fort Hancock 1490OtayMesa 1295 Presidio 1490
Tecate 1295 DelRio 1490Calexico 1295 EaglePass 1490Andrade 1295 Laredo 1490SanLuis 3763 FalconHeights 1490
Lukeville 3763 Roma 1490Sasabe 3763 RioGrandeCity 1808
Nogales 3763 LosEbanos 1808Naco 1780 Mission 1808
Douglas 1780 Hidalgo 1808Antelope Wells 1230 Pharr 1808
Columbus 1230 Progreso 1808SantaTeresa 1230 LosIndios 1808
ElPaso 1315 Brownsville 1808Fabens 1231
Operator Resilience Curves
Bribes Constant• Same results as shortest
path• At ten interdictions, only
decrease profit by $6000
Bribes vary by location• At three interdictions,
cannot move through NM• At ten, lost of $2 Million in
profits
Cannot use NM
Cannot use San Diego
cc
Bribes Vary by Location
Compound Strategy• Find a way to increase production cost and increase
distribution cost – Seize product– Force higher labor costs– Force him to develop innovative hiding and storage– Decrease of $20 Million Dollars in Profit
Demand Flow Model
Demand Flow Model
• Profit motivates flow • Realistic price-driven distribution• Delivery limited by capacity• Efficiency
– Demand profit per kilo • $19,000 (Austin)-$29,000(Dallas)
– Production cost per kilo• $10,000(Ciudad Obregon) - $15,500 (Saltillo)
• De la Vega limits flow through a border to 1000 kilos• Interdictions reduce flow on an arc to zero
Ciudad Victoria
END
START
MIDDLE
Demand vs. Supply Interdiction
Essentially no effect!
Stop Production in Saltillo
Reduce Production in Monterrey
MAXIMUM EFFECTS!
Border Interdictions
Supply Interdiction
What We Can Do?
• Account for randomness in the model– 20 model runs to find the expected value
• Influence De La Vega to forgo 2800 kilos in production capacity• Reduce amount delivered by 2700 kilos• Reduce profits by more than $15,000,000 with no additional resources
applied
IMPROVE CAPABILITY TO REDUCE EFFICIENCY!
Conclusion
Conclusions• Understand the network
– How do narcotics flow across the border?• Through every possible avenue by means of lowest cost
– What motivates flow?• Profit!
• How to best attack the network– Attack Demand
• Education, deterrence– Attack Distribution
• Best non-kinetic solution• Training (any improvement shows effect)• Improve border interdiction capability
– Attack Supply• Best kinetic solution• Cut the head off the snake
Further Analysis
• Consider Blue Force costs• Incorporate open desert crossing• Consider political/economic consequences
QUESTIONS?