BUSTING DRUGS WHILE PAYING WITH CRIME: The domestic cost of America’s drug control policy in foreign countries By Horace Bartilow Kihong Eom The University.

BUSTING DRUGS WHILE PAYING WITH CRIME: The domestic cost of

America’s drug control policy in

foreign countries

By

Horace Bartilow

Kihong Eom

The University of KentuckyDepartment of Political Science

Abstract

The existing literature on America’s drug control policy towards drug producing countries in Latin America contend that US policy has helped to create a political environment that has facilitated a significant rise in narco related violent crime, acts of terror against the state, and has destabilized the region in ways that foster the growth of the drug trade. But what effect does America’s drug control policy; specifically, drug interdiction and the immobilization of drug traffickers have on narco related violent crime in Central American and Caribbean countries who serve as strategic drug transit zones in the global drug industry? We answer this question by empirically analyzing drug crime data for Caribbean and Central American countries from 1984 to 2000 using two different estimation procedures. The first is a time series cross-section model and the second is a simultaneous equation model. After controlling for variables that affect overall crime the results show that increasing levels of drug interdiction and the immobilization of drug traffickers contributes to increasing levels of drug related crime in the Caribbean basin. These results suggest that US drug control policy may have the unintended effect of destabilizing the socio-political systems of Central American and Caribbean countries – important allies in America’s drug war and could ultimately discourage governments in the region from future cooperation with core aspects of US drug control policy.

The Relationship between Drug Trafficking and Violent Crime

In the major drug consuming countries in the world such as North America and Western Europe, it is now conventional practice to refer to a tripartite classification scheme of psychopharmacological effects, economic compulsive drives, and systematic violence to identify the relationship of drugs to violent crimes against people and property (Goldstein 1985; Tullis 1995).

The Relationship between Drug trafficking and Violent Crime (contd)

The psychopharmacological dimension relates to people becoming irrational, agitated, impulsive, uncontrollably anger and physically abusive even to the point of committing murder. The economic compulsive dimension is associated with violent criminal acts to obtain money for personal drug consumption (e.g. through burglaries and robberies).

The Relationship between Drug trafficking and Violent Crime (contd)

The systematic dimension relates to drug cartels, gangs, narco-insurgents and traffickers protecting their product or turf from law-enforcement officials, or from each other by resorting to extreme forms of violent behavior. The psychopharmacological and economic compulsive dimensions relates to violent crimes by drug abusers while the systematic dimension relates to violent crimes by cartels and traffickers.

American Drug Control Policy

Since violence is an integral part of the drug industry’s operation, American policy makers contend that US drug control policies; specifically, the interdiction of illicit drugs and the immobilization or arrest of drug traffickers is an effective means of reducing drug related violence in the global war on drugs (Reuter and Kleinman 1986).

American Drug Control Policy (contd)

American drug control policy is informed by the logic that the reduction of the supply of drugs will increase market prices in the US and as a result US consumers will therefore forgo drug consumption and drug abuse thereby reduce the likelihood of violent crime. By this logic, efforts to reduce domestic drug consumption in the US have led American policy makers to seek cooperation and collaboration with foreign countries in the areas of interdiction and trafficker immobilization (Toro, 1992).

American Drug Control Policy (contd)

To ensure foreign country cooperation in the ‘war on drugs’, the Reagan administration increased diplomatic pressure on drug producing and drug transit countries and strengthened the capacity of the US to impose the extraterritoriality of its criminal laws throughout the Western hemisphere. As a result, the administration introduced the Anti-Drug Abuse Act of 1986 and the priority of narcotics issues was, for the first time, placed on the international agenda. The US government developed a ‘certification’ practice to evaluate the performance of foreign country cooperation in the drug war.

American Drug Control Policy (contd)

The Anti-Drug Abuse Act made US financial assistance, positive votes within multilateral lending institutions and trade preferences conditional on foreign governments’ cooperation against drug trafficking (Perl, 1989; United States Senate, 1988b).

Existing Research and US Drug Control Policy

Students of US drug control policy towards drug producing countries in Latin America have consistently noted that US policies have increased the level of drug related violent crime in the region. These scholars argue that by implementing US drug policies in the region, American policy makers have frequently introduced US military forces into countries like Bolivia, Columbia and Peru to battle traffickers at the source of their operations.

Existing Research and US Drug Control Policy (contd)

For example Operation Blast Furnace - the intervention of U.S forces to destroy cocaine laboratories in Bolivia and the introduction of US military personnel to battle Narco-insurgents in Columbia and Peru. Consequently, American policy has militarized the region and facilitated an escalation of the drug related violence – primarily kidnappings, assassinations and the spread of narco-insurgent violent confrontations against governments in the region (Bagley 1992; Crandall 2002; Labrousse and Laniel 2001; Lupsha 1996; Tokatlian 1994).

The Limits of Existing Research

The problem with this line of argument is that the causal inference - namely, that US drug control policy has facilitated the escalation of drug violence in Latin America - is not convincing. The counterfactual is that even in the absence of US drug control policies, drug related violence would most likely increase due to the fact that illicit drugs are largely produced in Latin American countries and traffickers would naturally utilize violence against each other and against governments in the region in order to expand and protect their enterprise.

The Limits of Existing Research (contd)

Essentially, existing studies are unable to separate the level of drug violence that is endemic to countries that produce drugs from the reported enabling affects of US policies towards the region. A more systematic approach would be to estimate the impact of US drug control policies on drug violence in Central American and Caribbean countries who serve as transit zones and not drug producers, and where drug cartels like Cali and Medaine or narco-insurgents groups like FRAPH in Columbia and the Shining Path in Peru are not indigenous to the political landscape of these countries. In this way much of the drug related violence that is endogenous in existing country specific studies can be avoided.

The Central Argument

1. The only way in which US drug control policies - specifically drug interdiction – could effectively reduce crime is if we make the assumption that the demand for drugs is elastic. However, given the psychopharmacological addictive nature of drug abuse, we argue that the demand for drugs is inelastic. And therefore drug interdiction under conditions of demand inelasticity will increase drug related violence by drug users (the economic compulsive dimension) and by drug traffickers (the systematic dimension).

1 2 3 4 5 6 7 8 9

1

2

3

4

5

6

7

8

9

P2 = 5.5

Q2

S2

S2

P1 = 3.5

D

D

S1

S1

Q1

Figure. 1.The Effects of Drug Interdiction on Drug Price

when Demand is Elastic

1 2 3 4 5 6 7 8 9

1

2

3

4

5

6

7

8

9

P2 = 7

Q2 Q1

S2

S2

P1 = 4

D

DS1

S1

Figure. 2.The Effects of Drug Interdiction on Drug Price

when Demand is Inelastic

The Central Argument (contd)

2.The Immobilization of drug traffickers will most likely increase drug related violence due to the fact that traffickers will be replaced at a higher rate than the level of arrests. In fact the immobilization of drug traffickers may be counterproductive to the goal of reducing crime. Given the strong financial rewards of the drug industry, the removal of one trafficker merely opens up opportunities for another to enter. And new recruited traffickers are likely to commit more violence as a way of consolidating their position as they move deeper into the industry (Spellman, 1994; Spellman, 2000).

The Central Dependent Variable

The Drug Crime variable measures the volume of drug related crime (homicides and robberies) per 1000 persons for the Caribbean and Central American countries in our data set. The data is collected from various issues of the International Crime Statistics published by INTERPOL.

Central Explanatory Variables

1. Drug interdiction is measured in terms of the interdiction of cocaine and marijuana and is calculated by dividing a country’s yearly seizures of cocaine and marijuana in kilograms by the yearly number of a country’s active law enforcement personnel. The drug interdiction data was collected from various issues of the International Narcotics Control Strategy Report and from the Organization of American States (OAS): the Inter-American Drug Abuse Control Commission (CICAD). The law enforcement personnel data was adopted from various issues of The Military Balance (Institute for Strategic Studies, 1984 - 2000).

Central Explanatory Variables (contd)

2. Trafficker immobilization measures the yearly number of drug arrest of people convicted of drug trafficking. This data was also collected from various issues of the International Narcotics Control Strategy Report and from the Organization of American States (OAS): the Inter-American Drug Abuse Control Commission (CICAD).

Confounding Variable Number of Drug Users

We control for a host of variables that theoretically predict general crime and drug crime in particular. These include: the number of drug users in the Caribbean basin. Since the psychopharmacological and economic compulsive dimensions of drug related crimes relates to violent crimes committed by drug abusers – its is expected that increasing numbers of drug users is expected to increase drug related crime (Goldstein 1985; Tullis 1995). This data was also collected from various issues of the International Narcotics Control Strategy Report.

Confounding Variable Distance from the US

Drug trafficking is as much about the covering distance as it is about acquiring wealth. Since the Caribbean basin serves as a major transit point for drug traffickers, it is expected that countries that are geographically closer to the US will be more exposed to drug trafficking than those that are further away and will therefore have higher incidence of drug related violent crime. Our distance variable was collected from Direct-Line Distances, U.S Edition (Gary L. Fitzpatrick and Marilyn J. Modlin, 1986)

Confounding VariablesUnemployment

The rate of unemployment is traditionally used as a control variable in many criminological studies of crime. It is argued that high levels of unemployment increase the level of crime in general and drug related violent crime in particular,since the unemployed will be drawn to the lucrative drug industry (Ehrlich, 1973; Chaiken and Chaiken, 1982; Spelman, 2000). This variable was collected from the World Bank’s World Development Indicators 2001.

Confounding VariablesEconomic Growth

Economic growth is also a conventional control variable used in many criminological studies of crime. High levels of economic growth is expected to reduce general crime and drug related crime in particular. (Bruce D. Johnson, Andrew Gloub and Eloise Dunlap, 2000). This variable was collected from the World Bank’s World Development Indicators 2001.

Confounding VariablesTotal External Debt

High levels of external debt encourage people in developing countries to export illicit drugs to rich drug consumers in the developed countries (George,1992). Since indebtedness decrease personal disposable income – drug consumption and drug related violent crime will decrease as well. Therefore it is expected that the level of a countries total indebtedness will have a negative impact on drug violence. This variable was collected from the World Bank’s World Development Indicators 2001.

Confounding VariablesUS Drug Consumption/per street Value

It is important to control for the effects of US drug consumption on drug crime in the Caribbean basin. Since Central American and Caribbean countries serve as drug transit zones for markets in North America, increases in US drug consumption drives up profits for trafficking gangs and increase the likelihood for systematic violent crime between them. This variable is calculated by dividing the yearly consumption of drugs in the US by the average street price. This data was collected from the 2000 issue of the Office of National Drug Control Policy.

Estimation Model 1: Assumes an Exogenous Relationship between Drug

Control Policy and Drug Crime We analyze drug crime data for Caribbean and

Central American countries from 1984 to 2000 via the following time series cross-section equation:

Drug Crimeit = 0 + 1Drug Seizureit + 2Drug Arrestit + 3Drug Usersit + 4Distanceit + 5Unemploymentit + 6Economic growthit + 7External Debtit + 8Illicit Drug consumption per priceit + it

The Estimation method is OLS with panel corrected standard errors assuming V(it) = 2

i and COV(it, jt) ≠ 0 where i ≠j.

Model 2: Assumes an Endogenous Relationship between Drug Control Policy

and Drug Crime Our argument is that drug interdiction increases drug crime.

However, it is plausible that increases in drug crime would force governments to interdict drugs. And since replacement theory predicts that the immobilization or arrest of traffickers increase drug crime, its is also possible that increases in drug crime would lead to the arrest of drug traffickers. In other words, drug interdiction (cocaine and marijuana seizures) and the immobilization of drug traffickers are endogenously related to drug crime. Instrumental variables are generated to predict drug interdiction (cocaine and marijuana seizures) and drug arrest. These instruments are placed into the statistical model that predict Drug Crime via the following Three Stage Least Least Square Structural Equation:

Model 2: Three Stage Least Square Structural Equation Model of Drug Crime

1) Drug Crime1it = 1

0 + 11Marijuana Seizureit + 1

2Cocaine Seizureit + 1

3Drug Arrestit + 14Drug Usersit + 1

5Distanceit + 1

6Unemploymentit + 17Economic growthit + 1

8External Debtit + 1

9 US Drug consumption per priceit + 1it

2) Drug Arrest2it = 2

0 + 21Drug Arrestit-1 + 2

2Drug Crimeit + 2

3Marijuana Seizureit + 24Cocaine Seizureit +

25Distanceit + 2

6Drug Usersit + 27Unemploymentit +

28Economic growthit + 2

9External Debtit + 210 US Drug

consumption per priceit + 211 US Marijuana Consumptionit

+ 212Money Launderingit + 2

it

Model 2: Structural Equation Model of Drug Crime (Contd)

3) Marijuana Seizure3it = 3

0 + 31Marijuana Seizureit-1 + 3

2Drug

Crimeit + 33Drug Arrestit + 3

4Neighboring States’ Effortit + 35US

Bilateral Drug Interdiction it + 36Total Coast Line (Km) of CB

Countries it + 37CB Countries Distance from the US it + 3

8South

American Drug Seizures it + 39Per-capita purchasing power parity

in CB Countries it + 310US per-capita purchasing power parity it +

311Government Corruption in CB Countries it + 3

12Domestic

Illicit Drug Consumption in CB Countries it + 313Domestic Illicit

Drug Consumption in the US it + 314Economic Trade Openness of

CB Countries it + 315Individual Political Freedom and Civil

Liberties in CB Countries it + 316Institutional Political Openness

in CB Countries it + 317US Aid to CB Countriesit + 3

it

Model 2: Structural Equation Model of Drug Crime (Contd)

4) Cocaine Seizure4it = 4

0 + 41Cocaine Seizureit-1 + 4

2Drug

Crimeit + 43Drug Arrestit + 4

4Neighboring States’ Effortit + 45US

Bilateral Drug Interdiction it + 46Total Coast Line (Km) of CB

Countries it + 47CB Countries Distance from the US it + 4

8South

American Drug Seizures it + 49Per-capita purchasing power parity in

CB Countries it + 410US per-capita purchasing power parity it +

411Government Corruption in CB Countries it + 4

12Domestic Illicit

Drug Consumption in CB Countries it + 413Domestic Illicit Drug

Consumption in the US it + 414Economic Trade Openness of CB

Countries it + 415Individual Political Freedom and Civil Liberties in

CB Countries it + 416Institutional Political Openness in CB Countries

it + 417US Aid to CB Countriesit + 4

it

Model 2: Structural Equation Model of Drug Crime (Contd)

Where superscript indicates equation number, subscript i refer to a country and subscript t refers to a year. The estimation method is three stage least squares assuming COV (j

it, kit) ≠ 0 where j ≠ k.

RESULTS