Report No. 37820 Crime, Violence, and Development: Trends, Costs, and Policy Options in the Caribbean March 2007 A Joint Report by the United Nations Office on Drugs and Crime and the Latin America and the Caribbean Region of the World Bank ACRONYMS AND ABBREVIATIONS ADR Alternative Dispute Resolution CEM Country Economic Memorandum CFATF Caribbean Financial Action Task Force CGNAA COSAT Guard for the Netherlands Antilles and Aruba CONANI Consejo Nacional de la Niñez CPI Corruption Perceptions Index CPTED Crime Prevention Through Environmental Design CTS Crime Trends Surveys ­ United Nations DALYs Disability-Adjusted Life Years DHS Department of Homeland Security EBA Educación Básica para Adultos y Jóvenes ECLAC Economic Commission for Latin America and the Caribbean ENHOGAR Encuesta Nacional de Hogares de Propósitos Múltiples EU/LAC European Union/Latin American and the Caribbean FARC Fuerzas Armadas Revolucionarias de Colombia GDP Gross Domestic Product ICS Investment Climate Survey ICVS International Crime Victims Survey LAC Latin America and Caribbean OECS Organization of Eastern Caribbean States PATH Program for Appropriate Technology in Health RNN Royal Navy of the Netherlands RSS Regional Security System RTFCS Regional Task Force on Crime and Security UNODC United Nations Office on Drugs and Crime WDR World Development Report WHO World Health Organization Vice President: Pamela Cox Country Director: Caroline Anstey PREM Director: Ernesto May Sector Manager: Jaime Saavedra Chanduvi Lead Economist: Antonella Bassani Task Managers: Andrew R. Morrison Bernice Van Bronkhorst Gabriel Demombynes Theodore Leggett (UNODC) TABLE OF CONTENTS Preface _____________________________________________________________________ Acknowledgments Executive Summary __________________________________________________________i Chapter 1. Conventional Crime: an Overview___________________________________ 1 The Diversity of the Caribbean_________________________________________________ 2 How to Measure Crime? ______________________________________________________ 4 Homicide and Assault ________________________________________________________ 7 Violence Against Women ____________________________________________________ 12 Property Crime_____________________________________________________________ 13 Chapter 2. Organized Crime _________________________________________________ 15 Drug Trafficking ___________________________________________________________ 15 Other Forms of Organized Crime: Kidnapping, Money Laundering, Corruption________ 21 Chapter 3. Risk Factors for Crime and Violence_________________________________ 26 Conceptual Foundations of Risk Factors for Crime and Violence____________________ 26 Macro Analysis of Risk Factors for Crime in the World and the Caribbean____________ 27 Risk Factors for Crime at the Household Level: Evidence from Three Caribbean Countries________________________________________________ 34 Policy Implications ____________________________________________________ 40 Chapter 4. Socioeconomic Costs of Crime ______________________________________ 41 Approaches to Measuring the Costs of Crime and Violence ________________________ 42 Specific Costs of Crime______________________________________________________ 43 Total Costs: Estimates Using an Accounting Approach ___________________________ 50 Disability-Adjusted Life Years Lost to Violence _________________________________ 53 The Impact of Victimization on Self-Reported Life Satisfaction_____________________ 56 The Impact of Violent Crime on Economic Growth_______________________________ 57 Policy Implications _________________________________________________________ 59 Chapter 5. Youth Violence in the Caribbean: A Case Study of the Dominican Republic_______________________________________________ 61 Scope of Youth Crime and Violence ___________________________________________ 62 Factors Contributing to Youth Violence ________________________________________ 67 Policy Recommendations for Strengthening the Response to Youth Violence__________ 73 Conclusions _______________________________________________________________ 79 Chapter 6. Case Study: Criminal Deportations and Jamaica ______________________ 81 The Scale of the Expatriate Population _________________________________________ 83 Criminal Deportation________________________________________________________ 84 Policy Implications _________________________________________________________ 91 Chapter 7. Case Study: Drug Trafficking and the Netherlands Antilles______________ 93 Drug Trafficking on Commercial Air Flights: an Innovative Policy Response _________ 96 Maritime Trafficking_______________________________________________________ 101 Policy Implications ________________________________________________________ 102 Chapter 8. The Contributions of Criminal Justice Systems to the Control of Crime and Violence: A Case Study of Jamaica and the Dominican Republic ___________ 105 Introduction: Criminal Justice Reform in Civil and Common Law System ___________ 105 Country Cases: Two Approaches to Criminal Justice Reform______________________ 108 A Quantitative Evaluation of System Performance: Performance Statistics and Indicators _________________________________________________________ 112 Measuring the System's Workload: Reported Offenses __________________________ 113 Institutional Performance Measures ___________________________________________ 114 Impact Indicators: Using Crime and Arrest Data for Crime Prevention______________ 122 Conclusion: From Organizational Reform and Performance Measurement to Interagency Governance _________________________________________________ 126 Chapter 9. Guns and Crime: A Case Study of Trinidad and Tobago ______________ 128 Sources __________________________________________________________________ 130 Registration of Firearms and Seizures of Illegal Weapons_________________________ 131 Demand__________________________________________________________________ 132 Victims, Perpetrators and their Environment____________________________________ 132 Drug Trafficking and Gun-Related Criminality _________________________________ 133 Policy Implications ________________________________________________________ 136 Chapter 10. Public Policy of Crime and Violence Prevention: Regional and National Approaches ________________________________________________ 141 Sectoral and Cross-Sectoral Approaches _______________________________________ 141 Regional Initiatives to Address Crime and Violence _____________________________ 145 Priority Steps to Achieve Reductions in Crime and Violence in the Caribbean ________ 149 References _______________________________________________________________ 141 Annexes _________________________________________________________________ 175 FIGURES Figure 1.1: Ranking of Caribbean Countries in the Human Development Index -------------------------------- 3 Figure 1.2: Population Distribution of the Independent Caribbean------------------------------------------------ 3 Figure 1.3: Murder Rates by Region----------------------------------------------------------------------------------- 4 Figure 1.4: Murders per 100,000 Population by Year, Anguilla --------------------------------------------------- 6 Figure 1.5: Number of Murders in Guyana, Police and Public Health Sources---------------------------------- 7 Figure 1.6: Deaths Rates from Violence in Caribbean and Comparison Countries----------------------------- 9 Figure 1.7: Homicides per 100,000 in Guyana and Jamaica-------------------------------------------------------10 Figure 1.8: Homicides Rates in the Dominican Republic, St. Lucia and Trinidad and Tobago: 1999-2005 11 Figure 1.9: Assault Rates in Caribbean and Comparison Countries----------------------------------------------11 Figure 1.10: Rape Rates in Caribbean and Comparison Countries -----------------------------------------------12 Figure 2.1: Cannabis Eradication in Jamaica ------------------------------------------------------------------------17 Figure 2.2: Shares of Cocaine Flowing to the United States by Transport Corridors --------------------------17 Figure 2.3: Origin of Heroin Seized in the United States ----------------------------------------------------------19 Figure 2.4: Kidnappings per 100,000 Population in Trinidad and Tobago --------------------------------------23 Figure 3.1: Ecological Model for Understanding Crime and Violence ------------------------------------------27 Figure 3.2: Cross-Country Correlates of Crime ---------------------------------------------------------------------29 Figure 3.3: Robbery Rates vs. Mean Consumption by Neighborhood -------------------------------------------36 Figure 3.4: Distribution of Victimization by Crime and Quintile in Haiti---------------------------------------37 Figure 4.1: Responses to Fear of Crime in Dominican Republic -What Do People Stop Doing Due to Fear? ------------------------------------------------------------------------------------------------44 Figure 4.2: Responses to Fear of Crime in Haiti: Where Do People Stop Going Due to Fear?---------------45 Figure 4.3: Impact of Crime on Various Business Practices in Jamaica -----------------------------------------47 Figure 4.4: Crime Protection Measures Taken by Firms in Jamaica ---------------------------------------------48 Figure 4.5: Private Security Costs for Firms by Size of Enterprise as Percentage of Firm Revenue in Jamaica---------------------------------------------------------------------------------------------------49 Figure 4.6: Crime Protection Measures Taken by Firms in the Dominican Republic--------------------------49 Figure 4.7: Impact of Higher Crime on Businesses in the Dominican Republic--------------------------------50 Figure 4.8: How Reliable is WHO Homicide Information for the Caribbean? Homicide Rates According to WHO vs. Official Sources Homicide Deaths per 100,000 Population -------------54 Figure 4.9: Disability-Adjusted Life Years Lost to Violence in the Caribbean, 2002--------------------------55 Figure 4.10: Disability-Adjusted Life Years Lost to Violence vs. Other Causes Caribbean 2002 -----------55 Figure 4.11 Potential Boost to Annual Economic Growth Rate from Reducing Homicide Rate to Costa Rica Level-------------------------------------------------------------------------------------------59 Figure 5.1: Homicide Deaths by Age (10-29) in the Dominican Republic (2000-2005)-----------------------65 Figure 5.2: Trends of Minor Arrest over Past Decade in the Dominican Republic-----------------------------66 Figure 6.1: Number of Deportees and Number of Murders in Jamaica ------------------------------------------82 Figure 6.2: Caribbean-Born Populations in the United States-----------------------------------------------------84 Figure 6.3: Total Criminal Deportations from the U.S. to the Caribbean----------------------------------------85 Figure 6.4: Criminal Deportees from the U.S. per 100,000 Population of Home Country --------------------86 Figure 6.5: Total Criminal Deportations to Jamaica: 1998-2004 -------------------------------------------------86 Figure 6.6: Drug Deportations to Jamaica from the United States and the United Kingdom -----------------87 Figure 6.7: Breakdown of Criminal Deportees by Crime Type, All Three Source Countries, 2001-2004 --88 Figure 6.8: Age on Arrival in the U.S. of Jamaican Criminal Deportees (Percent of Headley's Sample)- 889 Figure 6.9: Age on Deportation from the U.S. of Jamaican Criminal Deportees (Percent of Headley's Sample) -----------------------------------------------------------------------------------------------------90 Figure 7.1: Kilograms of Cocaine Seized in Caribbean Territories in 2004-------------------------------------95 Figure 7.2: Kilos of Cocaine Seized in the Netherlands Antilles, 1980-2004-----------------------------------96 Figure 7.3: Couriers Detected arriving at Schiphol from Curacao, by quarter ----------------------------------99 Figure 7.4: Cocaine Seizures in Trinidad and Tobago ----------------------------------------------------------- 103 Figure 8.1: Sentenced and Unsentenced Prison Inmates, Dominican Republic, May 2003 to May 2006 - 119 Figure 8.2: Prison Deaths, Escapes, and Recaptures, Dominican Republic ----------------------------------- 120 Figure 8.3: Unnatural Deaths and Violent Incidents in Jamaican Prisons, 2000-2005----------------------- 121 Figure 9.1: Distribution of Injury Deaths by Type of Injury: Trinidad and Tobago, 1999-2003 ----------- 129 Figure 9.2: Police Reports of Narcotics Possession and Murders, and Certified Firearm Homicides: Trinidad and Tobago, 1992-2005---------------------------------------------------------------------- 133 BOXES Box 1.1 Crime Definitions---------------------------------------------------------------------------------------------- 5 Box 2.1: Financial Fraud in the Dominican Republic ­ The Baninter Case-------------------------------------25 Box 3.1: Haiti's Entrepreneurs of Violence -------------------------------------------------------------------------38 Box 5.1: The Sophisticated Extortion of Santo Domingo Gangs and Drug Lords------------------------------67 Box 8.1: Using Crime Data and Analysis to Formulate a Multi-Sectoral Crime Prevention Strategy: The Barrio Seguro Program in Capotillo ------------------------------------------------------------- 124 Box 9.1: Young Men, Drugs, and Guns---------------------------------------------------------------------------- 136 Box 9.2: UN Resolution and Plan of Action on Small Arms and Light Weapons---------------------------- 139 Box 9.3: Gun Buybacks ---------------------------------------------------------------------------------------------- 140 Box 10.1. Crime and Violence Prevention Components in Bank-Financed Integrated Slum Upgrading Operations ­ Jamaica Inner Cities Basic Services for the Poor Project--------------------------- 146 Box 10.2. The Pride in Gonzales Initiative, Port of Spain, Trinidad and Tobago: A Community-Driven Social Development Approach------------------------------------------------------------------------- 147 Box 10.3. The Need for Better Data on Crime -------------------------------------------------------------------- 151 TABLES Table 2.1: Highest Total Annual Cocaine Seizure, 1998-2004----------------------------------------------------21 Table 3.1: Cross-Country Analysis - Basic Regression Results---------------------------------------------------32 Table 3.2: Summary of Micro-Analysis of Risk Factors for Criminal Victimization --------------------------35 Table 4.1: Effects of Lifetime Physical Violence by Intimate Partner in Haiti ---------------------------------46 Table 4.2: The Costs of Crime in Jamaica: an Accounting Exercise ---------------------------------------------52 Table 4.3: Cross-Country Regression Estimate of the Effect of Violent Crime on Economic Growth ------58 Table 5.1 Youth as Victims: Homicide Rates in Select Countries------------------------------------------------64 Table 5.2: Highlights of Youth in Numbers in Latin America and the Caribbean------------------------------69 Table 5.3: Violence Prevention Strategies by Developmental Stage and Ecological Context ----------------76 Table 8.1: Homicide Clearance Rates in Jamaica by "Murder Motive," 2005 and 2006 -------------------- 115 Table 9.1: Murders Committed in Trinidad and Tobago Using a Firearm, 2001-2006 ---------------------- 129 Table 9.2: Woundings Committed in Trinidad and Tobago Utilizing a Firearm, 2000-2005 --------------- 130 Table 9.3: Firearm Seizures in Trinidad and Tobago------------------------------------------------------------- 132 Table 9.4: Modified Haddon Matrix Applied to the Analysis of Gun-Related Criminality (GRC) -------- 135 Table 9.5. Possible Interventions Specific to Guns and Criminality-------------------------------------------- 138 Table 10.1. Public Policy Approaches and Interventions to Address Urban Violence----------------------- 144 ANNEXES Annex 1.1. Caribbean Leaders on Crime--------------------------------------------------------------------------- 175 Annex 1.2. Tourism and the Caribbean Economy ---------------------------------------------------------------- 175 Annex 1.3. Emigration and Crime ---------------------------------------------------------------------------------- 175 Annex 1.4. Caribbean Police Forces and Prison Populations---------------------------------------------------- 176 Annex 2.1. The Jamaica Cannabis Trade -------------------------------------------------------------------------- 177 Annex 2.2. Patterns of Drug Transshipment by Organized Criminal Groups --------------------------------- 177 Annex 2.3. Dates of large Seizures of Cocaine Entering Canada from Caribbean Countries 2004 -------- 178 Annex 2.4. Organized Crime in Jamaica and the Dominican Republic ---------------------------------------- 178 Annex 2.5. Jamaat-al-Muslimeen in Trinidad and Tobago ------------------------------------------------------ 179 Annex 3.1. Caribbean Homicide Data Sources-------------------------------------------------------------------- 179 Annex 4.1. Impact of Crime Victimization on Satisfaction with Life------------------------------------------ 184 Annex 5.1. Risk Factors and Policy Responses to Youth Violence in the Dominican Republic------------ 185 Annex 5.2. Inventory of Governmental Initiatives for Youth Violence Prevention in the Dominican Republic------------------------------------------------------------------------------------ 188 Annex 5.3. Inventory of Civil Society Initiatives for Youth Violence Prevention in the Dominican Republic------------------------------------------------------------------------------------ 193 Annex 6.1. Criminal Deportations to Jamaica by Year, Offence, and Sending Country--------------------- 199 PREFACE This report is the result of a fruitful collaboration between the United Nations Office on Drugs and Crime (UNODC) and the World Bank. It addresses the issue of crime and violence in the Caribbean, a complex problem that prejudices both the social and economic development of the region. The Governments of the Caribbean countries recognize the seriousness of the problem and are exploring innovative policy responses at both the national and regional levels. Civil society organizations are doing their part as well by designing and implementing violence prevention programs targeting youth violence, violence against women, and other important forms of violence. Much, however, remains to be done. Some of the factors that make the Caribbean most vulnerable to crime and violence--the drug trade and trafficking of weapons are two important examples--require a response that transcends national and even regional boundaries. Also, promising initiatives at the national level must be evaluated for effectiveness so that scarce resources can be efficiently invested. This report is offered as a contribution to the ongoing dialogue in the region on approaches to address crime and violence. It is not intended to provide a definitive blueprint for action, but rather is offered as a tool to engage stakeholders--governments, civil society organizations, citizens, and international partners--in a serious dialogue on crime and violence, based on evidence and good practices from inside and outside the region. Francis Maertens Caroline Anstey Director, Division for Policy Country Director for the Caribbean Analysis and Public Affairs World Bank UNODC ACKNOWLEDGMENTS This report is a joint product of the United National Office of Drugs and Crime (UNODC) and the World Bank. The multisectoral team was led by Theodore Leggett (UNODC), Bernice van Bronkhorst (LCSFU), Gabriel Demombynes (LCSPP), and Andrew Morrison (PRMGE). Individual chapters were prepared by Mr. Leggett (overview chapters, criminal deportees in Jamaica, narcotics trafficking in the Netherlands Antilles), Mr. Demombynes (risk factors and costs), Linda McGinnis and Peter Holland (HDNCY: youth in the Dominican Republic), and Ms. van Bronkhorst and Mr. Morrison (public policy). The criminal justice chapter was written by Linn Hammergren (LCSPS) and Stephanie Ann Kuttner (SDV), based on a background paper on criminal justice prepared by Todd Foglesong (Vera Institute of Justice) and Christopher Stone (Harvard University). The firearms chapter was written by Bernice van Bronkhorst, Andrew Morrison, and Theodore Leggett, based on a background paper prepared by Yvette Holder (independent consultant) and Folade Mutota (Women's Institute for Alternative Development, Trinidad and Tobago). The team benefited from significant contributions from Ana Maria Diaz (LCSPP) and Mariel Fiat (LCCDO), as well as the excellent assistance of Ane Perez Orsi de Castro (LCSPP) in the preparation of the final report. The team would also like to thank the following individuals for their contributions: Lisa Bhansali, Teresa Genta-Fons, Lillian Mallet Crawford Abbensetts, Willy Egset, Dorte Verner, and Christina Malmberg Calvo. The team is very grateful for the work and insightful comments of its External Advisory Committee, chaired by Professor Anthony Harriott (Professor, University of the West Indies-Mona, Jamaica) and comprised of Dr. Franklin Almeyda (Minister of the Interior and Police, Dominican Republic), Mr. George de Peana (General Secretary, Caribbean Congress of Labor, Barbados), Professor Ramesh Deosaran (Director, Center for Criminology and Criminal Justice, University of the West Indies-St. Augustine, Trinidad and Tobago), Dr. Len Ishmael (Director General, OECS Secretariat, St. Lucia), Colonel Trevor MacMillan (Chairman, Standing Committee on National Security of the Private Sector Organization of Jamaica), the Honorable Dale D. Marshall (Attorney General, Barbados), and the Honorable Gail Teixeira (Special Advisor to the President and Director, Governance Unit, Office of the Presidency, Guyana). This task was guided and supervised by Jaime Saavedra (Sector Manager, LCSPR) and Antonella Bassani (Lead Economist, LCSPR). The peer reviewers were Pablo Fajnzylber (LCSFR), Wendy Cunningham (LCSHS), and Anthony Harriott (University of the West Indies, Mona Campus). EXECUTIVE SUMMARY In his 2006 New Year's address as then prime minister of Jamaica, P.J. Patterson said, "Without a doubt, the high level of violent crime remains our most troubling and pressing problem." In opening the Parliament of Trinidad and Tobago in September 2005, President George Maxwell Richards said the country was in crisis due to the escalating crime rate. Through multiple channels, crime and violence threaten the welfare of Caribbean citizens. Beyond the direct effect on victims, crime and violence inflict widespread costs, generating a climate of fear for all citizens and diminishing economic growth. Crime and violence present one of the paramount challenges to development in the Caribbean. Several factors which cut across the diverse countries of the region heighten their vulnerability to crime and violence. Primary among these is the region's vulnerability to drug trafficking. Wedged between the world's source of cocaine to the south and its primary consumer markets to the north, the Caribbean is the transit point for a torrent of narcotics, with a street value that exceeds the value of the entire legal economy. Compounding their difficulties, Caribbean countries have large coastlines and territorial waters and many have weak criminal justice systems that are easily overwhelmed. Key messages and recommendations from the report include the following: 1) Crime and violence are a development issue. The high rates of crime and violence in the region have both direct effects on human welfare in the short-run and longer run effects on economic growth and social development. Estimates suggest that were Jamaica and Haiti to reduce their rates of homicide to the level of Costa Rica, each country would see an increase in its growth rate of 5.4 percent annually. 2) While levels of crime and associated circumstances vary by country, the strongest explanation for the relatively high rates of crime and violence rates in the region--and their apparent rise in recent years--is narcotics trafficking. The drug trade drives crime in a number of ways: through violence tied to trafficking, by normalizing illegal behavior, by diverting criminal justice resources from other activities, by provoking property crime related to addiction, by contributing to the widespread availability of firearms, and by undermining and corrupting societal institutions. At the same time, it should be recognized that there is a trade-off between resources spent on combating drug trafficking and those spent on other forms of crime and violence prevention. 3) In general, there has been an over-reliance on the criminal justice approach to crime reduction in the region, to the detriment of other complementary approaches which can be effective in reducing certain types of crime and violence. Over the last few years, however, several countries such as Jamaica and the Dominican Republic are increasingly investing in different approaches. Crime prevention through environmental design (CPTED), the study and design of environments to encourage desirable behavior and discourage antisocial behavior, has significant potential to generate rapid decreases in property crime and some i forms of inter-personal violence. Integrated citizen security approaches have seen initial success in the Dominican Republic and should be explored elsewhere. These programs, by combining modern methods of policing with prevention interventions undertaken by both government and non-governmental organizations, are extremely promising. The public health approach, which focuses on modifying risk factors for violent conduct, is especially promising for addressing violence against women and youth violence. 4) At the same time, it is also crucial to note that certain types of crime and violence--in particular, organized crime and drug trafficking--are largely impervious to prevention approaches; a criminal justice-focused approach is essential in dealing with them. Within the criminal justice approach, there is much room for improvement. An especially urgent priority is the development of management information systems and performance indicators for better problem diagnosis, tracking of system outputs, monitoring reform programs and providing increased accountability to citizens. 5) These different approaches mean that there are multiple possible entry points to engage in violence and crime prevention. In one instance, the most promising approach may be in the context of a slum-upgrading project; in another, in the context of a reform of the health service; in a third, in the context of a reform of the criminal justice system. There is no one "ideal" approach. The common denominator is that successful interventions are evidence-based, starting with a clear diagnostic about types of violence and risk factors, and ending with a careful evaluation of the intervention's impact which will inform future actions. 6) Evidence from Jamaica and other countries shows that the average deportee is not involved in criminal activity, but a minority may be causing serious problems, both by direct involvement in crime and by providing a perverse role model for youth. More services should be offered to reintegrate deportees, along the lines of those provided by the Office for the Resettlement of Deportees in St. Kitts and Nevis. Options should be explored for deporting countries to shoulder a significant portion of the costs of these programs, in exchange for serious monitoring and evaluation of program impacts. 7) Given that Caribbean countries are transit and not producer countries of cocaine, interdiction needs to be complemented by other strategies outside the region (principally demand reduction in consumer countries and eradication and/or alternative development in producer countries). Within the region, policies should focus on limiting the availability of firearms and on providing meaningful alternatives to youth. Since the Caribbean nations have limited resources to effectively fight the drug trade, significant assistance should come from the destination countries in support of interdiction efforts. The case study of the Netherlands Antilles shows this to be both effective and in the self-interest of developed countries. 8) Gun ownership is an outgrowth of the drug trade and, in some countries, of politics and associated garrison communities. Within these environments, which ii promote the demand for weapons, reducing gun ownership is a difficult undertaking. Better gun registries, marking and tracking can help, as can improved gun interdiction in ports. Long run and sustained reduction in the demand for guns, however, will hinge on progress in combating drugs. 9) To address issues of youth violence, policy makers in the short run should borrow from the toolkit of evidence-based programs from other regions, such as early childhood development and mentoring programs, interventions to increase retention of high-risk youth in secondary schools, and opening schools after-hours and on weekends to offer youth attractive activities to occupy their free time. While there are a multitude of programs in the region that address youth violence, few if any have been subject to rigorous impact evaluation. In the medium and long run, impact evaluations should systematically document what works in youth violence prevention in the Caribbean. 10) This report has culled many different sources of data to present as comprehensive a picture as possible of crime and violence in the Caribbean. Yet it is clear that there are major data gaps that hinder policy making. Chief among them is the lack of regular, periodic victimization surveys that permit comparison of crime levels both across countries and over time. ROAD MAP OF THE REPORT The report is organized as follows. It begins with an overview of crime in the region, separately considering conventional and organized crime. Two subsequent chapters examine risk factors and the costs of crime for the region as a whole. Next, a series of chapters presents case studies designed to highlight particular issues in specific countries. These case studies were chosen in order to provide a detailed analysis of the most pressing issues that are amenable to policy making at the regional and national levels. The specific issues were chosen in consultation with stakeholders in the region to ensure that the report was responding to their demands and needs. The report ends with a chapter on public policy responses to crime in the region.1 Overview of Crime Trends Murder rates in the Caribbean--at 30 per 100,000 population annually--are higher than for any other region of the world and have risen in recent years for many of the region's countries. Assault rates, at least based on assaults reported to police, are also significantly above the world average. These reported rates are highly sensitive to the level of trust in the local police in general and the willingness to report domestic violence, in particular. Victimization surveys are needed to even approximate true levels of assault, yet standardized victimization surveys have rarely been undertaken in the Caribbean. 1Note that this report does not contain an in-depth analysis of political violence in the Caribbean. iii Murder Rates in Selected Caribbean Countries: 1999-2005 Murder Rates by Region of the World 35 Dominican Republic Caribbean 30 South/West Africa 29 30 Trinidad and Tobago South America 26 s St. Lucia East/SEAsia 22 25 denti Central America 22 res East Europe 17 20 000 Central Asia 9 100, East Africa 8 15 per North America 7 South Asia 4 urders 10 M Southeast Europe 3 Oceania 3 5 West/Central Europe 2 North Africa 1 0 Middle East /SW Asia 1 1999 2000 2001 2002 2003 2004 2005 Sources: (left figure) Procuraduría General of the Dominican Republic; Central Statistical Office of Trinidad and Tobago; Central Statistical Office of St. Lucia. Sources: (right figure) UN Crime Trends Survey and Interpol, 2002 or most recent year. Note: Figures shown are annual murder rates per 100,000 population. Violence against women affects a significant percentage of women and girls in the Caribbean. Police statistics offer only a very imperfect picture of violence against women, since the majority of these incidents are not reported to police and increased trust in police will increase reporting. To get a more precise idea of prevalence rates, one must use victimization surveys that focus on violence against women. One such regional victimization survey revealed that 48 percent of adolescent girls' sexual initiation was "forced" or "somewhat forced" in nine Caribbean countries (Halcon et al., 2003). According to the latest available data from the UNODC's Crime Trends Survey (CTS), which is based on police statistics, three of the top ten recorded rape rates in the world occur in the Caribbean. All countries in the Caribbean for which comparable data are available (Bahamas, St. Vincent and the Grenadines, Jamaica, St. Kitts and Nevis, Dominica, Barbados, and Trinidad and Tobago) experienced a rate of rape above the unweighted average of the 102 countries in the CTS. Despite their diversity, one thing all Caribbean countries have in common is that they have long been caught in the crossfire of international drug trafficking. The good news is that the flow of drugs through the region may be decreasing. The transshipment of cocaine to the United States, the most significant flow in economic terms, appears to be in decline. Cannabis production for export from Jamaica, the largest cannabis producer in the region, appears to be in a slump. iv Despite these recent shifts, large quantities of drugs continue to transit the Caribbean. In 2005, it is estimated that about 10 tons of cocaine transited through Jamaica, and 20 tons through Haiti and the Dominican Republic. In addition to drug trafficking, kidnapping and corruption are other forms of organized crime which affect the region. Two countries--Haiti and Trinidad and Tobago--have seen recent and rapid increases in kidnappings. The kidnapping rate nearly doubled in Trinidad and Tobago between 1999 and 2005 (Central Statistics Office of Trinidad and Tobago, 2006). Corruption is a difficult crime to measure. While there are methodological concerns about Transparency International's Corruption Perceptions Index (CPI), it remains the standard for international corruption comparisons and boasts one of the few datasets with near-global coverage. In the 2006 CPI, ten Caribbean countries were included in the world rankings. Haiti was ranked as the most corrupt country in the world, while Barbados was ranked as the 24th least corrupt country, ahead of many European countries. Risk Factors for Crime and Violence Potential risk factors for crime victimization encompass conditions at the individual, relationship, community, and societal levels. As a whole, Caribbean countries exhibit crime patterns similar to those in other countries. Both murder and robbery rates are higher in countries with low economic growth while murder rates are highest in countries and communities that are poor and have large populations of young men. But these factors alone cannot explain the high rates of crime in the Caribbean. In Caribbean countries overall, homicide rates are 34 percent higher and robbery rates are 26 percent higher than in countries with comparable macroeconomic conditions. Household-level victimization data from Jamaica, Haiti, and the Dominican Republic were analyzed to generate a profile of key risk factors. Poorer households in poor communities face higher risk of violent crime, while property crime more often strikes the wealthy. The presence of a large population of young men in the community is associated with higher levels of both types of crime in both Jamaica and the Dominican Republic. In all three countries, crime is highly concentrated in urban areas and in most cases in areas with high population density. While these risk factors provide a profile of who is most likely to be victimized, observable variables have low predictive power in a statistical sense because victims vary widely in their characteristics. In other words, crime victimization is a general phenomenon which strikes citizens of all stripes, and few if any are immune from the threat. This can be seen in part for Haiti by the fact that while property crimes most often strike those in the richest quintile, and those in the poorest quintile most often suffer injury due to crime, victims are spread across all five quintiles of the income distribution. v Murder Rates for Countries of the World vs. An important finding of this study GDP per Capita, Late 1990s is that in Jamaica a lower 0 percentage of crimes are reported to 10 the police in areas with higher ontilaup crime rates. The reporting rate can Haiti Jamaica plausibly be interpreted as a po 000 Dominican Republic measure of confidence in the police, Guyana 0,01reps Dominica as people will be more likely to 10 St.Kitts Trinidad & Tobago report when they trust the police Barbados Grenada Antigua & Barbuda deici and believe they will respond. Lack m of trust and confidence in the police Ho is then lower in areas with higher 1 local crime rates. This suggests also 1000 5000 10000 30000 that official police data distort the GDP per capita true geographic profile of crime, Sources: Own analysis. UNODC data and other (see Annex because official data are biased 3.1) for crime rates, Penn World Tables for GDP per capita. downwards for higher crime areas. Socioeconomic Costs of Crime and Violence This report reviews the literature on the effects of crime in the region and presents new analysis of the costs of crime in terms of loss of quality of life for victims, responses to fear of crime, and economic growth. Few studies in the region have examined socioeconomic costs of gender-based violence and its effects. Morrison and Orlando (2005) find that women victimized by physical violence in Haiti are: i) less likely to receive antenatal care; ii) more likely to suffer from genital sores and ulcers; and iii) more likely to be anemic (as are their children). Due to the high levels of violence in parts of urban Jamaica, residents are afraid to leave the homes and interact less often with friends and family who live elsewhere. Similarly, survey data from the Dominican Republic and Haiti show that people avoid activities and locations that are perceived to expose them to a high risk of criminal victimization. Another channel through which crime exacts costs is through its effects on businesses, which can be particularly damaging because they can involve both short-run costs and long-run consequences for development, by diverting resources to crime prevention measures and otherwise discouraging investment. In Jamaica, 39 percent of business managers in a World Bank survey responded that they were less likely to expand their business because of crime, and 37 percent reported that crime discourages investments that would improve productivity. Because of the key role that tourism plays in many Caribbean countries, the effects of crime on tourism are of particular concern. Alleyne and Boxil (2003) examined the relationship over time between tourist arrivals and crime in Jamaica and concluded that crime has discouraged tourists. vi It is possible to add up the total costs of crime through the accounting method. A study by Francis et al. (2003) found that the total costs of crime in Jamaica in 2001 came to J$12.4 billion, which was 3.7 percent of GDP. Security costs dominate the total costs of crime as calculated by the accounting method. Potential Boost to Annual Economic Growth The impact of crime on overall Rate from Reducing Homicide Rate to Costa economic growth can also be Rica Level estimated using cross-country panel th 6% data. Results from this kind of ow 5.4% 5.4% analysis suggest very large Gr e 5% potential gains from reduction in comnI violence for Haiti and Jamaica. atipaCreP 4% Both countries could boost annual 3% economic growth per capita by 5.4 alu 2% 1.8% 1.7% percent if they were to bring their Ann ni homicide rates down to the levels 1% se of Costa Rica. Guyana and the ea Incr0% Dominican Republic would also Dominican Guyana Haiti Jamaica benefit substantially, with potential Republic growth rate increases of 1.7 percent Source: Own analysis. and 1.8 percent, respectively. Criminal Deportees in Jamaica Each year, the United States, the United Kingdom, and Canada deport thousands of people convicted of various crimes to their countries of citizenship in the Caribbean. There is a widely held belief in the Caribbean that recent crime troubles can be tied directly to the activities of deportees who have learned criminal behavior in the developed countries. This report examines the situation in Jamaica, which is proportionately most affected by criminal deportations, and where officials have worried that the country's rising murder rate may be linked to the growing stock of deported convicts. The figure below shows both the number of murders and the number and source of deportee arrivals over the 1998-2004 period. Whether these two variables are related cannot be determined from this figure. vii Deportees and Murders in Jamaica by Year Some commentators have argued that many deportees left their home sn 8000 1500 countries at a young age and learned 7000 iotatropeDla Num criminal behavior while abroad. A 6000 ber recent study (Headley, 2005) 5000 1000 examined approximately 5000 4000 ofM records of criminals deported from 3000 u the U.S. between 1997 and 2003. The 2000 500 rder data indicate that the average age of a iminrC1000 s criminal deportee entering the United 0 0 8 9 0 1 2 3 4 States was 23 and the average age on deportation was 35. According to 199 199 200 200 200 200 200 official Jamaican statistics, 81 percent Canada deportees (left axis) were returned to Jamaica for non- UK deportees (left axis) violent offenses. While there is USA deportees (left axis) clearly heterogeneity in the pool of deportees, the typical deportee does Murders (right axis) not fit the profile of an individual who is likely to be a violent criminal Source: Social and Economic Survey of Jamaica; on return to Jamaica. Thus, it appears Jamaica Constabulary Force. unlikely that the average deportee is committing violent crime in Jamaica. At the same time, although the majority of the criminal deportees were deported for non-violent offenses, 224 convicted murderers were included in the flow between 2001 and 2004. Relative to the Jamaican population, this is not a small number, and it does not take a large number of offenders to have a potentially large impact. Are deportees contributing significantly to crime in other countries of the region? Of 332 criminal deportees returned to Barbados between 1994 and 2000, only 13 percent were subsequently charged with a criminal offense. Similarly, in Trinidad and Tobago, of the 565 deportees received between 1999 and 2001, only 15 percent were subsequently charged with a crime. This recidivism rate is low compared to the reoffense rate of prisoners released from local prisons. Developed countries provide a variety of forms of aid to the Caribbean to support development. Subsidizing reintegration for deported offenders would be a very cost- effective way of achieving similar ends. It would save Caribbean societies the cost and trauma of recidivism; reduce criminal justice costs involved in processing and incarcerating repeat offenders; and promote the stability essential to attracting investment, promoting tourism, and reducing emigration. Another result might be weakened international crime networks, which also will benefit developed countries. Guns in Trinidad and Tobago The data for several countries shows that not only have levels of crime and violence increased, but so too has the use of weapons in criminal acts. The profile of these viii incidents has also changed, with increased use of more powerful weapons resulting in higher mortality levels. The CARICOM Regional Task Force on Crime and Security recently commissioned a report on the proliferation of small arms and light weapons (SALW) in the Caribbean (CARICOM, 2002). The resulting report identified three levels of SALW proliferation in the region: countries with established high levels and patterns of armed crime (Jamaica), countries with emerging high levels of armed and organized criminality (Guyana, Trinidad and Tobago), and countries with indications of increased use and availability of small arms (Antigua and Barbuda, Barbados, Dominica, Grenada, St. Kitts and Nevis, Saint Lucia, St. Vincent and the Grenadines). At that time, it was determined that, among CARICOM nations, only Jamaica fell in the first category, with indications that military type weapons were available and that paramilitary units were operating (Burrows and Matthias, 2003). If such an evaluation were done today, Trinidad and Tobago might also be included in this tier, as the murder rate doubled between 2002 and 2005. In 2004, the country experienced 160 firearm murders, more than 450 firearm woundings, and 1,500 firearm incidents that did not result in injury (Trinidad and Tobago Police Service, n/d). A major factor contributing to the surge of guns-related criminality in the region is the trafficking of narcotics, which has facilitated the availability of firearms. More specifically, the firearms required for protection of contraband during transportation are smuggled in along with drugs. Within these environments that promote the demand for weapons, reducing gun ownership is a difficult undertaking. Better gun registries, marking, and tracking can help, as can improved gun interdiction in ports. In the long term progress will hinge on changes in the drug trade, changes in the "gun-culture," and progress in the implementation of international treaties and agreements on small arms and light weapons, such as the Protocol against the Illicit Manufacturing of and Trafficking in Firearms, their Parts and Components and Ammunition, supplementing the United Nations Convention against Transnational Organized Crime. Drug Trafficking and the Netherlands Antilles The drug trade is a prime driver of crime across the Caribbean. This report examines unusual interdiction efforts in the Netherlands Antilles, where authorities estimate that 75 percent of crime is drug-related. Some 60 percent of all the cocaine seized in the Caribbean in 2004 was seized in the Netherlands Antilles, and cocaine seizures increased dramatically between 2001 and 2004. Confronted with large numbers of people attempting to smuggle drugs by plane, authorities implemented a "100% Control" strategy, by which passengers landing in Schiphol Airport in Amsterdam and originating from the Dutch Caribbean, Suriname, and Venezuela are subject to extensive searches. Persons found with drugs on their person had the drugs confiscated and were immediately deported to their country of origin but were not arrested. Rather than attempting to scare off potential smugglers with ix the threat of incarceration, the Dutch approach was based on increasing the rate of interdiction to the point that smuggling became unprofitable. In other words, the focus was on the drugs, rather than the couriers. The authorities estimate that between 80 and 100 couriers per day were passing through the airport in 2003. This was cut to an estimated 10 a month by October 2005. While displacement effects would need to be tallied to properly evaluate the impact of this intervention, the 100% Control strategy has apparently disrupted what was once a major trafficking route. Commercial air flights are only one of the vectors through which cocaine transits the Netherlands Antilles and not necessarily the most significant one. The 100% Control approach has seized 7.5 tons of cocaine in 2.5 years, but multi-ton seizures can be made in a single instance of maritime interdiction. Detecting maritime trafficking requires intelligence work and a dedicated interdiction force. Even eliminating the Netherlands Antilles as a drug transit area altogether would not stop the flow of drugs, and displacement effects are inevitable. Nonetheless, focusing on the drugs rather than the couriers is a powerful approach because it defeats the "shotgun" technique, where traffickers are willing to sacrifice an endless stream of gullible couriers as long as sufficient quantities of drugs arrive for them to make a profit. It has proven itself to work in the case of the Netherlands Antilles, and could be tested in other contexts, including other Caribbean countries suffering from drug transshipment. The Antillean example also highlights the need for cooperation between Caribbean transshipment countries and destination countries in maritime interdiction. Youth Violence in the Dominican Republic Deaths and injuries from youth violence constitute a major public health, social and economic problem across the Caribbean, where youth are disproportionately represented in the ranks of both victims and perpetrators of crime and violence. Moreover, in many Caribbean countries violent crimes are being committed at younger ages. The Dominican Republic is one of the countries in which this pattern has become more stark over time, as rates of crime and violence overall have increased. In 2005, homicides of those aged 11- 30 accounted for approximately 46 percent of total homicide deaths. A wide variety of risk factors contribute to the prevalence of youth violence, including poverty, youth unemployment, large-scale migration to urban areas, drug trafficking, a weak education system, ineffective policing, the widespread availability of weapons, drug and alcohol use, and the presence of organized gangs. Nonetheless, youth violence is preventable. A broad range of strategies for preventing and reducing youth violence have been implemented in the Dominican Republic and elsewhere. Evidence from evaluations (unfortunately, almost exclusively in developed countries) documents that most highly effective programs combine components that address both individual risks and environmental conditions, by building individual skills x and competencies, supporting parental effectiveness, improving chances for youth to access and complete their secondary education, improving the social climate of schools, providing second chances, and promoting changes in involvement with peer groups. The best youth violence interventions target specific populations of young people associated with risk factors, such as school leavers, those involved with delinquent peers, gang members, and those exposed to family violence or substance abuse. Targeting high- violence communities with a holistic approach to address violence and emphasizing violence prevention directed at children and youth, as exemplified by the Barrios Seguros program in the Dominican Republic, is a very promising--albeit not yet formally evaluated--approach. Homicide Deaths by Age in the Dominican Republic, 2000-2005 100 90 80 2005 hsta 70 2002 defore 60 50 2000 40 mbuN 30 20 10 0 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Age Source: Dominican Republic National Police, unpublished data. Early child development interventions and effective parenting training for poor and at- risk children and their families are some of the most cost-effective investments in reducing youth violence and delinquency over time. Incentives for youths to complete their secondary education have also proven to be effective. Overall, reducing "mano dura" or repressive programs in favor of expanding prevention strategies (including prevention-focused law enforcement) would represent an effective and potentially cost-saving strategy. Finally, promoting interventions that reduce gun and alcohol availability--and their social acceptance--among youth can play a crucial role in countries like the Dominican Republic, where the use of both is widespread at young ages and the links to violence are significant. xi Criminal Justice Systems Many countries in the Caribbean have experimented with reform of their criminal justice systems, and the experience is mixed. This chapter focuses on the criminal justice reform experience of two countries: the Dominican Republic and Jamaica. Two important lessons emerge: (i) the need to pursue better coordination among institutions, including the introduction of information systems capable of tracking systemic performance and generating a set of performance indicators, and (ii) the desirability of linking criminal justice reforms to a broader, multi-sector strategy of crime and violence prevention. Performance indicators should not be the product of ad hoc efforts to measure outputs currently of interest. Instead, basic data on work processes should be routinely gathered and transferred to centralized databases, where it can be reviewed and analyzed for its broader implications. Once in place, such systems provide a very potent tool for reviewing organizational performance and identifying and diagnosing problems. They also facilitate the creation of new indicators as they are needed. A key performance indicator for the police is the "clearance" rate at which dockets leave the authority of the police and become the primary responsibility of the prosecuting authority. Another important indicator is the number and nature of complaints against the police. The most common indicators for measuring prosecutorial performance are the rate of convictions and the percentage or number of prisoners awaiting trial. Prison performance can be measured at a basic level by the rate of escape and amount of violence in prisons. The performance of the prison system in rehabilitating inmates can be measured by the re-arrest rate within a specified time period. Governments also need inter-agency governance mechanisms and incentives to help individual agencies in the justice sector align their work with system-wide goals. If sector-wide governance mechanisms are to succeed, their authority must go beyond performance measurement to include real executive powers (see Vera Institute of Justice, 2004)2. The recent history of justice reforms in both Jamaica and the Dominican Republic illustrates the need for such interagency mechanisms and effective performance evaluation systems. Public Policy of Crime and Violence Prevention: National and Regional Approaches In the Caribbean and most other regions, efforts to prevent violence have fallen into two categories: sector-specific approaches--such as criminal justice, public health, and conflict transformation and human rights--and cross-sectoral approaches--such as crime prevention through environmental design and citizen security. These approaches are complementary. For example, criminal justice reform initiatives (e.g. improved policing and better rehabilitation in prisons) can be pursued simultaneously with citizen security 2An example of sector-wide coordination currently limited to performance measurement is the role played by National Commission for the Coordination of the Reforms in Chile. xii programs that employ social prevention interventions and crime prevention through environmental design. In other words, there is no one "magic bullet" or single approach that can address all the risk and protective factors for crime and violence. It is important to note that many of the issues facing the Caribbean transcend national boundaries and require a coordinated regional response. Demand for drugs emanates from Europe and the United States; deportees are sent back to the region from the United States, the United Kingdom, and Canada; and many weapons that are trafficked are sourced from the United States. CARICOM has undertaken several important steps to deal with emerging regional security issues. Perhaps the most important regional initiative in the area of crime and violence reduction was the Regional Task Force on Crime and Security (RTFCS). The Task Force identified the following principal security threats to the region: illegal drugs, illegal firearms, corruption, rising crime against persons and property, criminal deportees, growing lawlessness, poverty and inequity, and terrorism. For the areas of illegal drugs, illegal firearms, terrorism, and deportees, the report included a short discussion of key issues and challenges and a long list of detailed recommendations at both national and regional levels (CARICOM, 2002).3 The Task Force presented its report to a meeting of Heads of Government in July 2002 and the recommendations of the Task Force were endorsed by this meeting. Caribbean heads of government endorsed a new Management Framework for Crime and Security in July 2005. This framework establishes a Council of Ministers responsible for security and law enforcement, a Policy Advisory Committee, and an Implementation Agency to implement CARICOM policy initiatives in this area. Yet no regional strategy in these areas can hope to succeed without significant support from OECD countries. This support has so far been limited predominantly to security sector reforms initiatives and drug interdiction, but has been lacking in the areas of weapons control and deportees. Nor does good policy making to reduce crime and violence happen by accident. The Regional Task Force on Crime and Security (CARICOM, 2002) called for the development of national crime control master plans and the establishment of national crime commissions to ensure multi-sectoral collaboration. National plans allow for cross- sectoral collaboration and serious discussions about the priorities in resource allocation. Equally importantly, they offer a vehicle for the involvement of civil society organizations, where much of the expertise in violence prevention resides. While the Task Force recommendations have been heeded in some countries, in some they have not. Preparation of a national plan should be a priority in these countries. Crime and violence are not immutable. While the Caribbean faces serious challenges, especially in the areas of drugs, guns and youth violence, intelligent policy making at the national and regional levels can make a difference. Given the high social and economic 3The key findings from these sections have been incorporated into the relevant sections of this report. xiii costs associated with crime and violence, the development of sound policies and programs is a key development priority for the region. The table below summarizes the key policy recommendations of the report. Summary of Policy Recommendations Priority Recommended Policy Actions Medium-Term Policy Actions · Create Injury Surveillance · Conduct regular, periodic and Systems standardized victimization surveys · Conduct impact evaluations that permit comparison of crime of all types of crime and levels both across countries and over violence prevention/ time reduction programs in the · Use data and analysis to identify Caribbean region Data/evidence geographical and demographic foci based policy- for interventions: making o Employ geographical information systems (GIS) to analyze crime trends and allocation prevention and control resources in large cities o Pilot "integrated citizen security" approaches to rapidly reduce crime in violent areas · Undertake institutional reform to promote systemic · Develop information systems and alignment and crime Criminal performance measurement indicators reduction Justice Reform to promote institutional efficiency · Modernize policing through and accountability the use of information systems and problem- oriented policing · Conduct impact evaluations Social and · Integrate crime and violence of social and situational Situational prevention into sectoral programs crime prevention initiatives Prevention such as slum-upgrading, education, · Scale up successful civil and health society crime and violence prevention programs xiv · Finance programs with proven track · Undertake impact record of success with youth violence evaluations that prevention systematically document · Invest in early childhood what works in youth development programs and programs violence prevention in the targeting children aged 4-10 Caribbean · Target specific youth-at-risk · Provide skills training and Youth Violence populations internships for at-risk youth · Reduce emphasis on ineffective "mano dura" programs · Apply existing laws separating incarcerated youth from adults · Restrict availability of alcohol and other drugs · Improve coordination with sending countries (who, when, share criminal · Enlist sending countries to records, etc.) finance programs for Deportees · Undertake robust research on reintegration of criminal contribution of deportees to crime deportees · Finance deportee reintegration, targeting deportees most likely to re- offend · Enlist consuming countries · Improve both commercial air and to provide financial and maritime interdiction by building on technical assistance for Drug regional successes such as drug- improved interdiction. Trafficking focused interdiction · Create alternative · Implement/expand drug abuse opportunities for youth. treatment programs · Reduce demand in consuming countries · Create/improve marking and tracking · Change gun culture systems · Implement and enforce · Create/improve national gun international and regional agreements on the Guns registries · Enforce gun laws and regulations for proliferation of Small Arms and Light Weapons licensing, selling, import/export (SALW) · Improve illegal gun interdiction xv · Create regional facilities for police training · Create regional · Follow-up and finance forensics/ballistic laboratory recommendations of the Regional and technical capacity Task Force on Crime and Security · Improve intelligence sharing · Create/strengthen national crime Regional and prevention plans and commissions international · Ratify the relevant international cooperation conventions on drug trafficking and organized crime, including in particular the United Nations Convention on Transnational Organized Crime and its protocols.. xvi 1. CONVENTIONAL CRIME: AN OVERVIEW The Caribbean countries are highly diverse in terms of their political structure, population size, and level of development. They share the experience of a colonial past and a geography which places them in the path of the international drug trade. According to figures from the United Nations Office of Drugs and Crime (UNODC) and Interpol, the overall Caribbean murder rate of 30 per 100,000 is higher than for any other region of the world. The region also has high levels of other violent crime, and violence against women is widespread. 1.1 Crime is arguably the number one social issue for much of the Caribbean, with its only serious rival being economic development (see World Bank, 2005a).4 Leaders are coming to realize, however, that the two issues are strongly related.5 Crime drives away investment, both foreign and domestic, and consequently slows growth. The Caribbean has been described as the most tourism-dependent area in the world, and crime is anathema for this industry.6 A second key industry--the financial services sector--is threatened by white collar crime and money laundering.7 The region is also one of the areas of the world most affected by brain drain, and there is evidence that crime is feeding this exodus in some areas.8 In these and many other ways, it is clear that crime is impeding the development of the Caribbean. 1.2 The Caribbean is especially vulnerable to crime for several reasons. It suffers from the disadvantage of being situated between the world's source of cocaine (the Andean region of South America) and its primary consumer markets (the United States and Europe) (UNODC, 2006).9 As small islands, Caribbean countries and territories have large coastlines and territorial waters to control relative to their ability to fund law enforcement coverage. Small criminal justice systems are easily overwhelmed in terms of police, courts, and prisons. Police must deal with seasonal tourist inflows, and, in some countries, the number of annual visitors actually exceeds the size of the local population. The Caribbean has some of the highest prisoner to population ratios in the world, and overcrowding interferes with the rehabilitation process.10 Finally, a number of countries 4 The open-ended question "What would you say is the single most serious challenge facing our country today?" was asked to a representative sample of the population of the islands of St Kitts and Nevis, St Lucia, St. Vincent and the Grenadines, and Antigua and Barbuda in Spring 2005. "The crime rate" was the most popular response, with 45 percent of respondents mentioning it as a first or second priority. The next most common response was "the rising cost of living", mentioned by 26 percent as a first or second priority. 5See Annex 1.1 for quotations from Caribbean leaders on the threat posed by crime. 6See Annex 1.2 for a discussion of the importance of tourism in Caribbean economy. 7See Box 2.1 for discussion of an example of white collar crime in the Dominican Republic. 8See Annex 1.3 for a discussion of the links between emigration and crime. 9In 2005, nearly all the world's cocaine came from Colombia (54 percent), Peru (30 percent), and Bolivia (16 percent). Most of this cocaine was used in North America (about half) and West/Central Europe (about a quarter). Cocaine prices in these countries were also generally higher than in other markets. 10See Annex 1.4 for discussions of Caribbean police forces and prison populations. 1 have experienced periods of political instability, which may have a long-term impact on crime.11 1.3 Discussions of crime typically distinguish between "organized crime" and "conventional crime." Organized crime typically refers to criminal enterprises (crime organizations that operate in similar fashion to businesses), while conventional crime includes all the common law offenses of murder, rape, assault, robbery, burglary, theft, fraud, and the like. This chapter discusses conventional crime, while the following chapter provides an overview of organized crime in the Caribbean. THE DIVERSITY OF THE CARIBBEAN 1.4 To those unfamiliar with the area, "the Caribbean" conjures rather uniform images of picture-postcard islands. The truth is that, on the contrary, it would be difficult to imagine a region that displays more diversity than the Caribbean. · The area was colonized by Denmark, France, the Netherlands, Spain, the United Kingdom, and the United States, and each power left its mark on the language, culture, and politics of the areas it held. Some countries retain varying degrees of dependence on these powers. · Some countries are highly developed, such as Barbados, which appears 31st in the world in the United Nations 2006 Human Development Index rankings, just above the Czech Republic. Others are extremely poor, such as Haiti, which ranks 154th, below both Kenya and Zimbabwe (see Figure 1.1) (UNDP, 2006). · Populations vary from about 4,000 (Montserrat) to over 11 million (Cuba). 1.5 This final factor, the uneven distribution of the region's population among countries, considerably complicates crime analysis. While there are 20 to 30 countries and territories in the Caribbean (depending on how they are counted) some 88 percent of the population is found in just five countries (see Figure 1.2). Since many islands are actually parts of larger countries (for example, the Netherlands Antilles and Aruba are part of the Kingdom of the Netherlands while Martinique, Guadeloupe, and French Guiana are French départments), they are often excluded from discussions of regional issues. If only fully independent countries are included in the analysis, Cuba comprises a third and the Dominican Republic and Haiti together half of the regional population. 11Key among these areas are Jamaica, which has experienced political violence since the 1940s, and Haiti, which has had repeated periods of instability since independence in 1804. 2 Figure 1.1: Ranking of Caribbean Countries in the Human Development Index Barbados 31 Cuba 50 St. Kitts and Nevis 51 Bahamas 52 Trinidad and Tobago 57 Antigua and Barbuda 59 Dominica 68 St. Lucia 71 Grenada 85 St. Vincent and the Grenadines 88 Suriname 89 Dominican Republic 94 Guyana 103 Jamaica 104 Haiti 154 0 20 40 60 80 100 120 140 160 Ranking Among All Countries of the Word in HDI Source: Human Development Report 2006. 1.6 Thus, the Caribbean is highly Figure 1.2: Population Distribution of the diverse, but the population of the Independent Caribbean Caribbean is concentrated in a few countries that may not represent the experiences of the smaller islands. For Cuba All others example, while the majority of the 32% 13% countries and territories are English- speaking, the bulk of the population is Spanish-speaking. Cuba comprises over a Jamaica quarter of the total regional population (a 7% third of the independent Caribbean), and its society and politics are unlike those of the other islands. Haiti makes up a fifth but is a unique case due its extreme Dominican Rep Haiti poverty and history of political instability 25% 23% and violence. Often both Cuba and Haiti are excluded from analysis due to data availability, but this also presents Source: UN Population Division. problems: it is difficult to speak of "the Caribbean" while excluding almost half the region's population, and doing so gives disproportionate weight to the experiences of the Dominican Republic and Jamaica. 1.7 For this reason, it is difficult to compare the Caribbean as a region to other regions in the world in terms of crime. Does the Caribbean have a serious crime problem? 3 It depends on which countries are included in the sample, and any generalization will gloss over a tremendous variety of experiences. But at least one international comparison, done by the UNODC based on standardized data sources, suggested that the Caribbean suffers from more murder per capita than any other region of the world (see Figure 1.3) (UNODC, 2005). Figure 1.3: Murder Rates by Region Caribbean 30 Southern and West Africa 29 South America 26 East and Southeast Asia 22 Central America 22 Eastern Europe 17 Central Asia 9 East Africa 8 North America 7 South Asia 4 Southeast Europe 3 Oceania 3 West and Central Europe 2 North Africa 1 Middle East and Southwest Asia 1 - 5 10 15 20 25 30 35 Murders per 100,000 population Source: UNODC Crime Trends Surveys and Interpol, 2002 or most recent year. HOW TO MEASURE CRIME? 1.8 In general, crime data are extremely problematic, and the Caribbean region provides an excellent case study of just how deceptive they can be. The best source of information on crime comes from household surveys, such as the standardized crime surveys conducted under the aegis of the International Crime Victims Surveys (ICVS). Unfortunately, only one country in the Caribbean has participated in the ICVS: Barbados. Information from other survey sources can be interesting, but rarely approaches the degree of precision needed for sound analysis of the crime situation. For example, the Latinobarómetro poll covers 18 countries in Latin America but only asks the most general questions about crime. 1.9 The official crime figures published by national governments are much more problematic. They are generally based on police statistics, and the police figures are 4 largely based on cases that are reported to the police by the public. Unreported cases cannot be recorded, and there is good reason to believe a great deal of crime is not reported in the region. Making comparisons across jurisdictions is even more complicated, because the precise rate of under-reporting varies between countries, and countries where the criminal justice system enjoys a good deal of public confidence tend to have higher rates of reporting. On the other hand, as Chapter 3 shows, it is precisely in the most crime ridden-areas that reporting rates are the lowest. 1.10 In addition, definitions of crime vary greatly between countries. Even for what seems like an easily defined offense such as murder, definitions vary widely, and crimes like burglary, robbery, and sexual offenses are defined very differently across jurisdictions. These differences are strongest when comparisons are made between entirely different legal traditions, and there are many in the Caribbean. In addition, the point in the criminal justice process when an alleged offense is recorded as a crime differs greatly between countries. This complicates comparison between civil law jurisdictions (such as in Spanish-speaking Caribbean countries) and common-law jurisdictions (such as the Anglophone states of the Caribbean). Box 1.1 Crime Definitions The following definitions of crime types are used by the United Nations Office on Drugs and Crime in the biennial Survey of Crime Trends and Operations of Criminal Justice Systems: · Intentional homicide is death deliberately inflicted on a person by another person, including infanticide. · Assault is physical attack against the body of another person, including battery but excluding indecent assault. · Rape is sexual intercourse without valid consent. · Robbery is the theft of property from a person, overcoming resistance by force or threat of force. · Theft is the removal of property without the property owner's consent. · Automobile theft is the removal of a motor vehicle without the consent of the owner of the vehicle. · Burglary is unlawful entry into someone else's premises with the intention to commit a crime. · Kidnapping is unlawfully detaining a person or persons against their will (or national equivalent, e.g. using force, threat, fraud or enticement) for the purpose of demanding for their liberation an illicit gain or any other economic gain or other material benefit, or in order to oblige someone to do or not to do something. 1.11 Another issue is the usefulness of crime rates in very small countries. Small countries typically stand out in the global rankings of a number of variables when these figures are calculated as a rate per 100,000 citizens, because a relatively small number of incidents can result in high rates. Many countries and territories in the Caribbean have populations of less than 100,000. Low populations also make interpretation of rates over time difficult, since trends tend to be erratic. The figure below showing murder rates in Anguilla (population 10,000-16,000, depending on source) illustrates this point. Since 1995 Anguilla has experienced either zero, one, or two murders per year--yet the 5 difference between zero and two murders is enough to move Anguilla from a low to a high murder rate (see Figure 1.4). Figure 1.4: Murders per 100,000 Population by Year, Anguilla 18 17 16 16 0,000 10 14 pers 12 10 10 der 10 9 8 mur 8 ded or 6 ecr 4 e 2 Polic 0 0 0 0 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: Anguilla Statistics Unit. 1.12 Crime rates are based on the ratio of crimes to full-time residents, but the actual population of small countries with large tourist industries can swell considerably during peak tourism periods. For example, the Bahamas, with a resident population of about 321,000, received nearly 1.5 million stay-over tourist arrivals in 2003 (World Travel and Tourism Council, 2004). Any of these people could have become a crime victim or perpetrator during their stay, so the ratio of crimes to the resident population can be deceptive. 1.13 To complicate matters further, there are frequently significant discrepancies in crime rates within a country depending on the source of the data. In the case of homicide rates, police and public health data sometimes differ substantially. The case of Guyana is instructive: two different murder counts are published in the same statistical bulletin-- one from the police and one from the ministry of health. This difference is not surprising since public health definitions differ from those used in the criminal justice system, but the ratio between the two varies also considerably over time. The public health figures show a clear increasing trend, while the police figures do not (see Figure 1.5). 6 Figure 1.5: Number of Murders in Guyana, Police and Public Health Sources 100 91 90 79 80 74 70 67 100,000 60 51 pers 50 40 33 der 30 Mur 20 10 0 1999 2000 2001 Police Ministry of Health Source: Guyana Bureau of Statistics, 2005. 1.14 UNODC attempts to overcome some of these difficulties through its biannual Survey on Crime Trends and the Operations of Criminal Justice Systems, or CTS. A questionnaire is sent to all United Nations Member States giving standardized definitions of each crime type and asking the respondents to fit their crime figures into the appropriate categories. Of course, this is a difficult exercise, and it is not clear how carefully each respondent complies, but it does provide a better basis for comparison than the figures published by the national police forces.12 In the discussion that follows, the CTS data is used, but this data must be treated with caution for all the reasons described above. HOMICIDE AND ASSAULT 1.15 Murder figures are generally considered the most reliable indicator of the violent crime situation in a country, since most murders come to the attention of the police, which is not the case with crimes like robbery and domestic violence. But definitions of murder still vary widely. In some countries, attempted murder is included, since legally this is deemed equivalent to the completed act, while in others "culpable" homicides (the killing of another through reckless behavior) are included. 12 Unfortunately, not all Member States supply this information, or they do so erratically, and so global coverage is far from complete. Obviously, countries with extreme stability problems are not able to provide reliable statistics, and it is precisely in these areas that the problems are likely to be worst. For example, little police data are available for Haiti, a country which is probably one of the more dangerous in the region. 7 1.16 For these reasons, it is best to use standardized data (such as those gathered by CTS or Interpol) for comparative analysis. Unfortunately, the data from both CTS and Interpol are far from complete in this region. Time series CTS figures for the Caribbean are limited to four countries and Interpol figures to two, and even these two data sets differ in many respects. 1.17 Aside from the police statistics, public health authorities maintain records of the numbers of murders in many countries, which are collected by the World Health Organization (WHO) and their regional affiliates. These public health data are available for a wider range of countries than are CTS data, including very small countries. Public health definitions of murder are generally more expansive than those of criminal law, so these figures will be greater than those reported to the UNODC and Interpol. 1.18 According to figures published by the WHO from 2002, a number of Caribbean countries record a rate of death due to violence13 in excess of the unweighted average of the 191 countries reporting worldwide, but far below the rates of countries like Colombia, Sierra Leone, or El Salvador (see Figure 1.6). These countries include countries with acknowledged violence issues, such as the Dominican Republic and Haiti, as well as some surprises, such as St. Kitts and Nevis and St. Vincent and the Grenadines. WHO data for Jamaica are clearly in error. According to WHO data Jamaica has one of the lowest rates of intentional violence in the world. According to the police statistics, however, the homicide rate was 56 per 100,000 residents in 2005--one of the highest rates in the world (Jamaican Constabulary Force, 2006). 1.19 Where no standardized data are available, national police statistics can give us a sense of trends within a country over time. The Jamaican Constabulary Force makes long-term comprehensive statistics available to the public, and the trend for Jamaica is quite remarkable (see Figure 1.7). After a peak during the politically-driven violence of the 1980 election, murder rates dropped to pre-election levels and were relatively stable throughout the 1980s. It was during this period of time that many Jamaican criminal groups ("posses") were active in the crack cocaine markets of the United States (Gunst, 1995). Ironically, as the crack trade declined, Jamaican murder rates went up. This may be due to former traffickers turning to income sources more directly rooted in violence, such as extortion. It may also be a symptom of the declining political control of the "area dons" in the "garrison" communities and the proliferation of a larger number of "corner dons," more likely to victimize their own constituencies (Mogensen, 2005), or it may also be related to the deportation of some Jamaican criminals that had been preying on the United States, the United Kingdom, and Canada (see Chapter 6). 1.20 Between 1982 and 1997, the homicide rate more than doubled in Jamaica; from 2003 to 2005 alone, the homicide rate rose from 36 to 58 per 100,000, before falling to 49 per 100,000 in 2006. Jamaica authorities attribute the 15% decline in the homicide rate in 2006 to law enforcement action. While the recent decline is encouraging, murder rates in Jamaica remain among the highest in the world, and similar rapid reductions in the past 13Violence in this context is defined as intentional violence minus suicides and war deaths. 8 have not abated a long-term upward trend. Still, more than 200 people are alive today that would not have been had the murder rate remained at 2005 levels. Figure 1.6: Deaths Rates from Violence in Caribbean and Comparison Countries Colombia 72.4 Bahamas 21.2 St. Vincent and the Grenadines 13.0 St. Kitts and Nevis 11.9 Antigua and Barbuda 10.2 Dominican Republic 10.2 Haiti 10.2 Guyana 9.9 Worldwide Average 9.0 Trinidad and Tobago 8.8 Barbados 8.5 St. Lucia 7.6 Cuba 5.3 Suriname 4.2 Japan 0.6 0 10 20 30 40 50 60 70 80 Deaths due to violence per 100,000 Source: WHO 2002. 1.21 The murder rate in Guyana (see Figure 1.7) has been a subject of frequent alarm in the country, but it displays no clear trend. An average of about 75 to 125 murders occurred annually since 1978, aside from brief peaks in 1991 and 2002-2003. The most recent increase has been attributed to gang warfare in the Buxton area (Guyana Government Information Agency, 2004). Given the small population (about three quarters of a million people), the murder rate is quite volatile and has varied from a low of about 10 per 100,000 in 1986 and 2000 to a high of 27 per 100,000 in 2003. As noted above, the police figures differ considerably from Ministry of Health figures, which were higher to varying degrees in 1999 (33 murders per 100,000 versus 12), 2000 (51 versus 10) and 2001 (67 versus 19). 1.22 Shorter time series on homicide rates are available for the Dominican Republic, St. Lucia, and Trinidad and Tobago (see Figure 1.8). All three countries show rapidly rising rates over the 1999-2005 period. In the Dominican Republic, the homicide rate almost doubled, from 14 to 27 per 100,000. In St. Lucia, the rate more than doubled, from 9 to 20, and in Trinidad and Tobago the rate more than quadrupled, from 7 to 30 per 100,000. 9 Figure 1.7: Homicides per 100,000 in Guyana and Jamaica 60 58 sredru 54 50 49 44 42 42 m 40 40 eddr 000, 37 37 36 3334 31 100 30 28 27 coer 25 26 26 erp 23 23 23 20 20 21 20 18 18 1918 191819 18 19 ce-iloP 17 17 17 17 15 16 16 1415 14 14 14 15 12 13 1213 12 121312 12 10 11 11 11 10 10 10 10 10 8 8 0 0 1970 1975 1980 1985 1990 1995 2000 2005 Year Guyana Jamaica Sources: Jamaica Constabulary Force, 2006; Guyana Bureau of Statistics, 2005. 1.23 According to the police statistics, the French départements (provinces) in the Caribbean also show remarkably high murder rates, despite their relative affluence. Out of 100 French départements, Guiana ranks first in terms of homicide, suffering a total of 94 murders in 2004, for a police recorded murder rate 51 murders 100,000.14 Guadeloupe and Martinique ranked third and fourth of 100, with 11 and 10 murders per 100,000 respectively (Direction Centrale de la Police Judiciaire, 2005). 1.24 In sum, homicide rates in the Caribbean are quite high by world standards. For the majority of countries for which time series data are available, homicide rates seem to be rising quite rapidly. Guyana is a notable exception to this trend. 1.25 While most murders make it to the attention of the police, lesser assaults are clearly highly reliant on reporting. The country's attitude toward domestic violence is pivotal, as is the level of trust in the local police. Figure 1.9 reports assault rates for countries with available CTS data. Even Dominica, with its 93 assaults annually per 100,000 citizens, is far in excess of the unweighted average of 10 for the 89 countries for which comparable data are available. The Bahamas reports the highest rate in the CTS dataset: 1,697 per 100,000, twice as high as the second highest rate recorded: 754 per 100,000 in Swaziland, a world leader in murder. This high rate is probably attributable to genuinely high levels of violence, possibly affected by the tourist influx, combined with high rates of reporting to a trusted police force. Victimization surveys are needed to even 14This is much higher than its closest rival, Southern Corsica, which registered 21 murders per 100,000. 10 approximate true levels of assault, yet standardized victimization surveys have rarely been undertaken in the Caribbean. Figure 1.8: Homicides Rates in the Dominican Republic, St. Lucia and Trinidad and Tobago: 1999-2005 35 30 30 27 0 25 00, 25 100 21 21 20 20 per 20 1918 15 15 urders 15 14 14 13 13 m 1212 10 9 9 7 recorded cei 5 olP 0 1999 2000 2001 2002 2003 2004 2005 Dominican Republic Trinidad and Tobago St. Lucia Sources: Procuraduría General de la República Dominicana; Central Statistical Office of Trinidad and Tobago; Central Statistical Office of St. Lucia. Figure 1.9: Assault Rates in Caribbean and Comparison Countries Bahamas 1697 Swaziland 754 Jamaica 219 Barbados 103 Dominica 93 Worldwide Average 10 Singapore 2.5 0 200 400 600 800 1000 1200 1400 1600 1800 Recorded Assaults per 100,000 population Source: Crime Trends Surveys ­ United Nations (various years). 11 VIOLENCE AGAINST WOMEN 1.26 Violence against women affects a significant percentage of women and girls in the Caribbean.15 Rape is greatly under-reported everywhere in the world, and relatively high recorded rape rates can actually be due to high levels of trust that reporting will result in positive outcomes for the victim. But given the fact that most of the rapes reported did occur, high rape rates do reflect a serious problem. According to the latest available CTS data, three of the top ten recorded rape rates occur in the Caribbean, including the top rated: the Bahamas. All countries in the Caribbean for which comparable data are available experienced a higher rate of rape than the unweighted average of 102 countries responding to the CTS: 15 rapes per 100,000. Figure 1.10: Rape Rates in Caribbean and Comparison Countries Bahamas 133 Swaziland 121 St. Vincent and the Grenadines 112 Jamaica 51 St. Kitts and Nevis 45 Dominica 34 Barbados 25 Trinidad and Tobago 18 Worldwide Average 15 Saudi Arabia 0 0 20 40 60 80 100 120 140 Reported Incidents of Rape per 100,000 Source: Crime Trends Surveys ­ United Nations (various years). 1.27 According to police records in the Dominican Republic, women between the ages of 15 and 34 account for nearly two-thirds of all violent deaths among women, despite only representing 36 percent of the female population (ALEPH, 2006.) Those at highest risk are young women working as domestic laborers and those having recently ended an intimate relationship. In approximately 63 percent of cases, the perpetrator is the victim's husband (boyfriend) or ex-husband (ex-boyfriend), followed by mothers (14 percent) and fathers (10 percent) (Caceres, F. and G. Estevez, 2004). 1.28 Police statistics offer only a very imperfect picture of violence against women, since the majority of these incidents are not reported to police. To get a more precise idea 15 Violence against women was defined by a declaration of the General Assembly of the United Nations (Declaration on the Elimination of Violence against Women, 1993) as "any act of gender-based violence that results in, or is likely to result in, physical, sexual or mental harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life." 12 of prevalence rates, one must use victimization surveys that focus on violence against women. Unfortunately, as is the case with other crimes, there are no victimization surveys using a common methodology that have been widely used across the Caribbean to document levels of violence against women (see Ellsberg and Heise, 2005).16 Consequently, we are left with isolated, country-specific data on prevalence rates; while this is undoubtedly better than relying on country-specific official crime reports, it does not allow cross-country comparisons. 1.29 One notable exception is a regional study carried out in nine Caribbean countries in 1997 and 1998 on adolescent health. According to this study, 48 percent of adolescent girls' sexual initiation was "forced" or "somewhat forced" in these nine countries (see Halcon et al, 2003).17 In Kingston, Jamaica in the early 1990s, 17 percent of 13 and 14 year-olds had been raped or been victims of attempted rape, and 33 percent had experienced unwanted physical contact or verbal enticements to have sex (Walker et al., 1994). 1.30 While there has been no comparable methodology used in a large number of Caribbean countries on domestic violence, national level prevalence studies generate surprisingly similar victimization rates. In Haiti, a recent DHS survey found that 28.8 percent of ever-married women had been beaten by a spouse. Older data from nationally representative surveys undertaken in Antigua and Barbuda and Barbados in 1990 indicated that 30 percent of all women in each country had been victimized by physical violence at the hands of an intimate partner at some point in their lives (see Heise et al, 1999).18 These lifetime prevalence rates are within the range of those found outside the region in a study recently undertaken by the World Health Organization. Rates of lifetime physical violence ranged from a low of 12.9 percent in urban Japan to a high of 61 percent in rural Peru, with a non-population weighted average prevalence rate of 36.3 percent over the 15 study sites in the ten countries (García-Moreno, C. et al., 2006). 1.31 In sum, violence against women seems to be endemic in Caribbean countries-- as it is in most countries around the world. PROPERTY CRIME 1.32 As the preceding sections have shown, the rate of murders is the single best crime statistic for comparative analysis. Data on property crime are far less reliable, since the rates of reporting the various offenses vary so greatly between jurisdictions and across time. In wealthier countries, a greater share of the population may have insurance and thus strong incentives for reporting victimization. These countries may appear to have much higher rates of property crime than poorer areas. Similarly, in some countries 16Standard victimization surveys do not do a good job of capturing violence against women. Special protocols are needed, both to protect the safety of interviewers and respondents, and to elicit accurate answers. 1732 percent of male adolescents' first intercourse was forced. The nine countries and territories were: Antigua and Barbuda, Bahamas, Barbados, British Virgin Islands, Dominica, Grenada, Guyana, Jamaica, and St. Lucia. 18In Antigua and Barbuda, the sample was of women ages 29-45, while in Barbados it was women 20-45. 13 reporting minor crimes is seen as a civic duty, even if the prospects for a positive outcome are poor. If there were good victim survey data indicating the level of under- reporting, the police figures could be interpreted in this light. Unfortunately, only Barbados has participated in the ICVS surveys. 1.33 Generally speaking, the best reported property crime is vehicular theft, and it is often used as an indicator of overall property crime levels. There are two problems with doing this in the Caribbean. First, the opportunities to successfully steal a car are limited on small islands, so vehicular theft rates may not be indicative of overall property crime levels. Second, since not everyone owns a car, the rate of property theft should properly be calculated as the number of incidents per motor vehicle in the country, not per 100,000 population. However, recent and reliable information on the number of vehicles in the Caribbean is not available. 1.34 The fact that reporting rates are key determinants in the levels of recorded crime is demonstrated by the available statistics on other forms of property crime. Australia leads the world in reported burglary rates, with Dominica scoring third, after Denmark (United Nations Crime Trends Survey, various years). While it is possible that some of the most developed countries, with low levels of inequality, suffer from extreme levels of burglary, it is more likely that this offense is simply under-reported in countries where it actually occurs more often. 1.35 Theft rates are subject to even higher levels of under-reporting. It is highly unlikely, for example, that Demark has an incidence of major theft 7,000 times higher than in Poland (3,449 episodes per 100,000 persons in Demark, versus 0.5 per 100,000 in Poland). Dominica has the second highest rate of recorded major theft among the 65 countries reporting this figure in the Crime Trends Survey, but the meaning of this distinction is highly dubious. 1.36 Given what is known about other aspects of crime in the region, it is highly unlikely that Dominica, which bills itself as "the nature island of the Caribbean" has 20 times the burglary rate and 30 times the theft rate of Jamaica, an island with acknowledged crime problems. 14 2. ORGANIZED CRIME Organized crime in the Caribbean is involved in a variety of activities, among which the most widespread is drug trafficking. In 2005, it is estimated that about 10 tons of cocaine transited through Jamaica, and 20 tons went through Haiti or the Dominican Republic. However, the flow of drugs through the region has decreased with the shift of trafficking to the Central American corridor. Cannabis production for export from Jamaica, the largest producer in the region, also appears to be in a slump. But demand from Europe may be increasing, and much of this traffic transits the Caribbean. Kidnapping and corruption also affect the region. Haiti and Trinidad and Tobago have seen recent and rapid increases in kidnappings. There appears to be wide variation in levels of corruption in the region. According to Transparency International's 2006 Corruption Perceptions Index, Haiti is perceived to be the country in the world with the highest levels of corruption, while Barbados was ranked among the least corrupt. 2.1 Violence in the Caribbean is aggravated by the presence of organized crime. There are no internationally comparable statistics on organized crime generally, but its activities can be detected in conventional crime statistics, as well as in seizures of contraband, particularly drugs. This information is supplemented by details emerging from criminal cases, as well as criminal intelligence, largely from developed countries.19 DRUG TRAFFICKING 2.2 Despite their diversity, one thing all Caribbean countries have in common is that they have long been caught in the crossfire of international drug trafficking. The Caribbean suffers greatly from the supply from the south and the demand in the north. Drug flows through a country can aggravate crime in a range of ways: · They produce local drug use problems, as couriers are often paid in product rather than cash and are compelled to sell on local markets. This has secondary effects on domestic crime problems, including youth gangs, prostitution, and market- related violent and property crime. · Drug transactions involve firearms, and firearms are often traded for drugs. · Movement of drugs inevitably involves corruption of local law enforcement officials, as well as other civil servants. · Laundering the proceeds of drug sales undermines legitimate economic activity. 19Since the United States is the main source of drug demand affecting the Caribbean, U.S. information sources are used extensively in this chapter. 15 2.3 The good news is that the flow of drugs through the region may be decreasing: · The transshipment of cocaine to the United States, the most significant flow in economic terms, has declined, as cocaine consumption in the United States dropped from its high levels in the 1980s (see National Institute on Drug Abuse, 2005).20 · Cannabis production for export from Jamaica, the largest cannabis producer in the region, appears to be in a slump (see Royal Canadian Mounted Police Criminal Intelligence Directorate, 2005).21 On the high end of the market, Jamaican cannabis faces competition from the growing popularity of indoor-produced sinsemilla in consumer countries in North America and Europe.22 Partly as a result, cannabis eradication rates dropped sharply after 1999, though they have recovered somewhat since then (see Figure 2.1).23 · The Caribbean has traditionally been the preferred drug transshipment corridor of Colombian organized crime groups, often working with wholesalers and retailers from the Dominican Republic. For a number of reasons (including effective law enforcement in the Caribbean, the break-up of the major Colombian cartels, and growing stability in Colombia), Mexican organized crime groups have supplanted the Colombian organizations throughout much of the United States. Mexican groups prefer to transship along the Central American coast and then across the U.S./Mexico border. They also prefer to market Mexican-produced cannabis on the low end of the market24 (see Figure 2.2 for the most commonly used transport corridors for the shipment of cocaine to the U.S.). 20While changes in the methodology of the national household surveys in the United States make long- term national comparisons problematic, cocaine use among U.S. high school 12th graders plummeted from 13.1 percent annual prevalence in 1985 to 5.1 percent in 2005. 21 The one area where the Caribbean continues to dominate is hash oil, a product Central and Atlantic Canada have long imported from Jamaica. Jamaican traffickers in Ontario are responsible for smuggling hash oil directly from the Caribbean and also manufacturing it within Canada. 22The share of sinsemilla in cannabis samples submitted to the Marijuana Potency Monitoring Project, which tests every seizure in which federal agents are involved, grew from 5 percent in 1985 to 31 percent in 2003. This sinsemilla averaged 14 percent THC in 2003, much higher than most field-grown cannabis. 23See Annex 2.1 for a discussion of the Jamaican cannabis trade. 24See Annex 2.2 for a discussion of patterns of drug transhipment by organized crime. 16 Figure 2.1: Cannabis Eradication in Jamaica 1000 894 833 900 811 743 ed 800 692 695 705 atcida 700 600 456 473 517 445 er 500 412 es 332 400 ar ct 300 He 200 80 100 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: International Narcotics Control Strategy Report, various years. Figure 2.2: Shares of Cocaine Flowing to the United States by Transport Corridors 11 3 1 2 1 1 100% 10 90% 12 26 27 22 80% 33 inea 43 70% 30 occ 60% of 50% 90 40% 88 72 72 77 30% 59 66 54 Share 20% 10% 0% 1998 1999 2000 2001 2002 2003 2004 2005 Central America/Mexico Caribbean Direct Source: U.S. National Drug Intelligence Center, various years. 2.4 There are, however, several contrary developments: · While cocaine use in the United States has declined, it is increasing in Europe. For example, cocaine use rates in Spain and England today are close to United States rates (see World Drug Report, 2006).25 Europol reports that 40 percent of the cocaine entering Europe transits the Caribbean.26 25In 2003, adult annual cocaine usage stood at 2.7 percent in Spain and 2.4 percent in England and Wales. In the USA, it stood at 2.8 percent in 2004. 26Europol, statements made at the Horizontal Drug Group (HDG) of the European Union, 10 January 2006. See Annex 2.3 for a discussion of cocaine transhipment patterns. 17 · Canada also poses an alternative market, with a high level of cocaine use.27 In 2004, large (in excess of 100 g) seizures of cocaine were made entering into Canada from a number of countries in the Caribbean, including Jamaica, Trinidad and Tobago, the Netherlands Antilles, Guyana, St. Vincent and the Grenadines, Haiti, St. Lucia, Grenada, Barbados, Suriname, and Dominica. · Since 1993, the United States has become increasingly dependent on Colombian heroin (see Figure 2.3). In terms of value per unit volume, heroin is worth more than cocaine, and thus it is still most commonly trafficked using couriers on commercial air flights. However, many large volume shipments have also been detected. · Synthetic drugs seem to be increasingly transiting or originating in the Caribbean. For example, on 28 November 2005, 259,338 ecstasy tablets were found when three abandoned bags were seized at the airport in Puerto Playa, believed to have originated in the Netherlands and to be destined for the United States.28 In 2005, an LSD lab was discovered on the Dutch side of St. Maarten, (U.S. Department of State, 2006) a relatively rare find for law enforcement anywhere in the world due to the difficulty of synthesizing this drug. 2.5 Thus, despite recent shifts and as indicated above, large quantities of drugs continue to transit the Caribbean. There are many reasons to expect that the Caribbean will continue to be a key drug transshipment area: · Colombia remains the source of most of the United States' cocaine and heroin, so the Caribbean will remain an attractive transshipment route. · The Caribbean has language, historical, commercial, and legal ties to some of the major consumer countries, including the massive tourism industry. These ties include daily direct air flights and container traffic. · The Caribbean diaspora, particularly nationals of Jamaica and the Dominican Republic, are key in domestic drug distribution in parts of North America and Europe.29 Expatriates from Europe and the United States living in the Caribbean further enhance this network.30 · The torrent of remittances from expatriate populations, as well as a large financial services industry, provides cover for money laundering. 27Annual use prevalence of 2.3 percent in 2004. See World Drug Report 2006 Volume 2 page 385. 28Personal communication, Col. Radhames Antigua Sanchez, Director CICC, Direccion Nacional del Control de Drogas, 1 February 2006. 29See Annex 2.4 for a discussion of organized crime in Jamaica and the Dominican Republic. 30Some arrest data reported to the UNODC in response to the Annual Reports Questionnaire gives an interesting perspective into the role citizens of the former colonial powers play in drug trafficking in this region. In Jamaica in 2004, for example, there were nearly as many foreigners arrested for drug offenses (221) as Jamaicans (234). The single largest group were British (although some may have been British of Caribbean ancestry). There were 68 British citizens arrested for cocaine offenses, compared to 79 Jamaicans. Overall, there were more foreigners arrested for cocaine offenses (124) than Jamaicans (79). Similarly, in Suriname in 2004, 269 Surinamese were arrested for drug offenses, compared to 174 Dutch nationals. 18 Figure 2.3: Origin of Heroin Seized in the United States 100% 90% ed 80% Seiz 70% oin 60% 50% Her 40% of e 30% 20% Shar 10% 0% 1990 1991 1992 1993 1994 1995 1998 1999 2000 2001 2002 South America Mexico Southeast Asia Southwest Asia Source: U.S. Drug Enforcement Administration Heroin Signature Program (National Drug Intelligence Center, 2002, 2003 and 2004; Drug Enforcement Administration, 1996) 2.6 It is not surprising, then, that of the 20 countries identified as major drug transit or major drug producing countries by the United States government in 2006, four are in the Caribbean: the Bahamas, the Dominican Republic, Haiti, and Jamaica.31 In commenting on this list, however, the United States government has noted the interdiction successes of the Bahamas and the Dominican Republic.32 In contrast, Haiti has received less praise for their cooperation with the United States, and has even faced being designated as a non-cooperating country.33 2.7 On the other side of Hispaniola, Dominican authorities estimate that, in 2004, most (67 percent) of the cocaine entering their country came by land, which means it came from Haiti. This is actually a substantial decrease from 2003, when the figure was 90 percent (UNODC, 2004). The Dominican Republic is seen as a "command, control, and communications" center for drug operations in the Caribbean. It is also used to store drugs, before onward shipment to Puerto Rico or the United States.34 31United States Presidential Determination No 2006-24, September 18, 2006. 32Statement of Acting Assistant Secretary of State for International Narcotics and Law Enforcement Affairs, Ambassador Nancy Powell, Briefing on Ongoing Diplomatic Activities at the UN and Other Current U.S. Foreign Policy Issues, Foreign Press Center, New York City September 15, 2005. 33The U.S. Drug Enforcement Agency has been involved in the arrest of many individuals linked to former-President Artistide's Lavalas party, including former senators and senior police officials, such as the former chief of presidential palace security and former chief of the anti-drug police. The fact that Haiti does not seize as much cocaine as many of her neighbours is more reflective of law enforcement capacity then drug flows: the Haitian anti-drug police has only about 40 members. See International Crisis Group, "Spoiling security in Haiti." Latin America/Caribbean Report No 13. Brussels: International Crisis Group, 2005. 34Statement of Rogelio E. Guevara, Chief of Operations of the U.S. Drug Enforcement Administration, before the House of Representatives Committee on International Relations, 10 October 2002. 19 2.8 The U.S. list does not include Puerto Rico, a U.S. territory. Puerto Rico has the third busiest port in North America, and its commonwealth status with the United States means that any shipments leaving Puerto Rico are not subject to search in the United States. In 2001, the head of the United States Drug Enforcement Administration described Puerto Rico as "an excellent gateway for drugs destined for East Coast [U.S.] cities."35 It has been the traditional transit area used by Dominican groups to bring cocaine into the United States.36 2.9 Cuba's relatively low ranking in terms of seizures has been attributed to recent enforcement efforts. According to the Nicaraguan police, seaborne cocaine has abandoned the routes along the Cuban coast due to increased patrolling, and now hugs the coastline of Nicaragua and Honduras.37 2.10 But there are few islands that cannot claim annual seizures of over half a ton of cocaine at some point in the recent past (Table 2.1). In addition, substantial amounts of heroin have been seized in the Dominican Republic, Aruba, the Netherlands Antilles, Bermuda, Puerto Rico, and, most recently, Trinidad and Tobago.38 The Dominican Republic in particular seems to be seeing an increasing flow of heroin, with seizures growing from seven kilos in 1998 to 122 in 2005. This is not surprising given the role Dominican expatriates have played in drug distribution in the United States. 35Statement of R. Marshall, head of the U.S. Drug Enforcement Administration, before the Senate Caucus on International Narcotics Control, 15 May 2001. Marshall also noted that while "Mexican trafficking groups normally charge Colombian traffickers 50 percent of each shipment to transport their product through Mexico, ...Dominican and Puerto Rican groups offer the same service for as low as 20 percent." 36Statement of Micheal T. Horn, chief of International Operations, U.S. Drug Enforcement Administration, before Senate Foreign Relations Committee, Subcommittee on the Western Hemisphere, Peace Corps, Narcotics, and Terrorism, 16 July 1997. 37Lt Col. Miguel Guilarte, head of the Nicaraguan anti-drugs unit of the border police, in LatinNews.com, "Cuba: Antinarcotics drive becomes a `national security issue.'" Latin American Regional Report: Caribbean and Central America, March 2005. 38Fifteenth Meeting of the Heads of National Drug Law Enforcement Agencies, Latin America and the Caribbean, Santa Marta, Colombia, 18- 21 October 2005 20 Table 2.1: Highest Total Annual Cocaine Seizure, 1998-2004 Highest annual seizure Country/territory Year total (kilograms of cocaine) 1998-2004 Puerto Rico 1998 10344 Netherlands Antilles 2003 9025 Bahamas 2003 4361 Jamaica 2002 3725 Guadeloupe 1998 3222 Cuba 1999 2444 Dominican Republic 1998 2342 British Virgin Islands 2001 2159 Turks and Caicos 1998 2075 Cayman Islands 1999 1926 Haiti 1998 1272 Martinique 2003 1138 Anguilla 2001 926 Trinidad and Tobago 2001 821 Aruba 1998 794 Antigua and Barbuda 2001 767 Bermuda 2001 667 Source: UNODC Data database 2.11 One possible side effect of drug transshipment is the development of local use problems, as have started to manifest themselves with heroin and ecstasy in the Dominican Republic.39 This is particularly true if the drug is moved through a diffuse network of couriers (as is typical with heroin) rather than through a small number of large shipments orchestrated by a few central players. For the most part, drug use levels in the Caribbean are lower than in the destination markets, suggesting that high levels of organization have typified the market in the past. Today, however, in a number of countries, drugs are dealt on the street by loosely organized groups of young people. OTHER FORMS OF ORGANIZED CRIME: KIDNAPPING, MONEY LAUNDERING, CORRUPTION 2.12 Other forms of organized crime affect the region as well. Three forms of important organized crime are kidnapping, money laundering and corruption.40 39 Unfortunately, the survey data on drug use in the Dominican Republic are rather dated. In 2001, heroin users comprised about 4 percent of the treatment population in the country, the highest share in Latin America except Mexico, where the drug is produced (5 percent). Authorities estimate that half of the ecstasy trafficked through the country is for local use. 40While there is some evidence that forms of human trafficking occur in the Caribbean, there is little data on the scale of this problem. There were only 56 convictions in the entire Western Hemisphere for human trafficking in 2004, up from 27 in 2003. Compare this to Europe and South Asia, both of which lodged over 1,200 convictions in 2004. See Office to Monitor and Combat Trafficking in Persons, 2005. 21 2.13 Official statistics on kidnapping are not reflective of the extent of a kidnapping problem in a country. There are several reasons for this. Definitions of kidnapping vary greatly, from the very specific offense of abducting a child for the purpose of extorting money out of parents to more generic laws proscribing moving people against their will. Most kidnapping laws do not distinguish between the unlawful removal of a person for non-financial motives (such as often occurs during a custody dispute over a child) and organized kidnapping for ransom. In some jurisdictions, kidnapping is often a supplemental charge brought in to strengthen a case that was really about something else, such as cases of abduction for the purposes of sexual assault. Perhaps most importantly, many kidnappings go unreported to the police, as this is often a demand made by the kidnappers as a condition for the safe return of the person. In fact, it is precisely in countries where kidnappings are common that reporting often declines, as professional negotiators enter the picture and supplant the role of official law enforcement. 2.14 Thus, it is more useful to look at trends within a country than to compare rates across countries. Using this metric, two countries in the Caribbean have seen rapid increases in their kidnapping rates in recent years. The kidnapping rate nearly doubled in Trinidad and Tobago between 1999 and 2005.41 And in Haiti, the acceleration between September and December 2005 was meteoric: according to 2005 figures compiled by MINUSTAH, 56 individuals were kidnapped in September, 63 in October, 74 in November and 241 in December--almost all in urban areas (United Nations, 2006). 2.15 Kidnapping is a relatively new problem for Trinidad and Tobago, possibly fuelled by wealth disparities, ethnic divisions, and the proximity of the country to Colombia. While suspicion typically falls on the Muslim extremist group Jamaat-al- Muslimeen,42 there is evidence that less organized criminals are tapping into this lucrative activity. The ransoms can be considerable: more than seven million U.S. dollars was paid out for the release of six members of the Greater Chaguanas Chamber of Industry after they were kidnapped in 2004 (Maharaj, 2004). The situation in Haiti may be even more dire, which is remarkable given that the crime really only emerged in the past two years. 2.16 Historically, the Caribbean has been susceptible to money laundering for a number of reasons. It has long focused on providing offshore financial services to the United States and European markets, but due to the small state capacity, did not have sufficient oversight resources. It is physically located along one of the world's premiere drug and cash thoroughfares. The tourism sector generates numerous cash-based businesses through which dirty money can flow undetected. The Dominican Republic provides an example of both of these techniques, as transport of bulk cash remains one of the primary means of transporting drug proceeds from the United States to 41These figures are based on information provided by the Central Statistical Office of Trinidad and Tobago. 42See Annex 2.5 for a description of Jammat-al-Muslimeen's activities in Trinidad and Tobago. 22 Figure 2.4: Kidnappings per 100,000 Population in Trinidad and Tobago 20 19 18 18 17 population 16 14 13 12 12 10 10 100,000 10 per 8 6 4 dnappingsik 2 of # 0 1999 2000 2001 2002 2003 2004 2005 Source: Elaborated from Central Statistical Office of Trinidad and Tobago the country. Local casinos and currency exchange houses are also said to be a major channel.43 Finally, the region is the recipient of a huge number of small cash remittances from the United States and Europe, constituting up to 30 percent of the GDPs of some Caribbean countries (ECLAC, 2005).44 2.17 Considerable progress has been made in combating money laundering in recent years, in part due to measures to block terrorist financing. The Financial Action Task Force (FATF) on Money Laundering, a Caribbean inter-governmental body, has published a list of "non-cooperative" countries and territories since 2000. Four Caribbean countries and one territory (Bahamas, Cayman Islands, Dominica, St. Kitts and Nevis and St. Vincent and the Grenadines) were placed on this list in 2000, and Grenada was added in 2001. By 2003, actions in all these countries and in the Cayman Islands led them to be removed from the non-cooperative list.45 The FATF has been a major factor in Caribbean countries improving their policies to combat money laundering (Sullivan, 2006). 2.18 The United States government also publishes a list of countries where money laundering is believed to be taking place. A number of Caribbean nations are currently listed as "jurisdictions of primary concern" for money laundering by the U.S, State Department: Antigua and Barbuda, the Bahamas, Belize, Cayman Islands, the Dominican 43Statement of Rogelio E. Guevara, Chief of Operations of the U.S. Drug Enforcement Administration, before the House of Representatives Committee on International Relations, 10 October 2002. 44It is estimated that Haiti's remittances are around US$1 billion, while its GDP is only about US$3 billion. After the collapse of its coffee and offshore assembly sectors, Haiti's top sources of foreign exchange are said to be remittances, foreign aid, and the drug trade. 45Bahamas and the Cayman Islands were removed in June 2001; St. Kitts and Nevis in June 2002; Dominica in October 2002; Grenada in February 2003; and St Vincent and the Grenadines in June 2003 (Sullivan, 2006). 23 Republic, Haiti, and St. Kitts and Nevis. It is important to note, however, that many major financial centers are also on the list, including the United States itself, the United Kingdom, Canada, France, Germany, Italy, and the Netherlands, as well as many other countries. 2.19 "Corruption" is a broad heading for a variety of activities related to the misuse of public office, including the extortion of bribes, procurement fraud and related kickback schemes, and the direct embezzlement of public funds. Not all of these activities are "organized" in the sense that large networks of individuals are concerned, but official corruption is extremely valuable to organized crime groups, and these groups play a strong role in promoting it, especially among law enforcement and border control officials. A case of corruption that had serious impacts on the economy of the Dominican Republic is described in Box 2.1. 2.20 The Caribbean may be vulnerable to corruption for a variety of reasons. In small states, the educated elite may be widely connected by blood, marriage, or social ties, and so it is difficult to distinguish nepotism from simple market necessity. Areas with a history of authoritarian government may also be susceptible, as the lack of transparency during these regimes can create a culture of corruption that endures even after the coming of democracy. 2.21 While there are serious methodological concerns about Transparency International's Corruption Perceptions Index (CPI), it remains the standard for international corruption comparisons, and boasts one of the few datasets with near-global coverage. In the 2006 CPI, Haiti ranked as the country perceived as most corrupt in the world (163rd out of 163 countries). Ranking for some other Caribbean countries were as follows: Guyana (121), Dominican Republic (99), Suriname (90), Trinidad and Tobago (79), Grenada (66), Jamaica (61), Dominica (53), and Barbados (24). 2.22 While Jamaica scores near the worldwide average in cross-country surveys of corruption, bribery and lack of transparency in government contracts are considered by Jamaicans to be important problems. Jamaica ranks poorly in perceptions of favoritism shown by government officials towards well-connected firms and individuals when deciding on policies and contracts. This is closely linked with the pressure exerted on businesses by protection rackets, and reflected in the high perceived costs imposed on businesses by organized crime (World Bank, 2004). 2.23 The best quantified information on corruption in the region comes from the Dominican Republic, a country that is included in the various cross-national surveys of Spanish-speaking countries, as well as Transparency International's Corruption Barometer; it was, in fact, the only Caribbean country to be included in the 2005 Corruption Barometer report. According to the Corruption Barometer 2005, some 16 percent of the Dominicans surveyed had paid a bribe in the previous year, the average amount being equivalent to US$274 when adjusted for purchasing power parity. More than half complained that they had to pay a bribe in order to access a public service to which they were entitled and to avoid problems with authorities (Transparency International, 2005). The World Bank's Investment Climate Survey showed that 72 24 percent of firms operating in the Dominican Republic reported "being affected" by corruption, and 21 percent of firms reported paying bribes in order to gain government contracts (World Bank, 2006a). 2.24 In the 2005 Latinobarómetro survey, 72 percent of Dominicans said that corruption had increased "a lot" in the previous three years, and 34 percent said they expected it to increase "a lot" in the next three years. Corruption among the police and political parties was emphasized. At the same time, 30 percent said that some progress had been made against corruption in the last year, but only 39 percent said that corruption would ever be solved (Latinobarómetro, 2005). Box 2.1: Financial Fraud in the Dominican Republic ­ The Baninter Case The Banco Intercontinental, S.A., better known as Baninter, began to experience liquidity problems late in 2002, due primarily to the alleged appropriation of funds by the bank's owners, and to a lesser extent because of an increase in deposit withdrawals. The bank had maintained a parallel bank with off-book operations through a system called "Interbanco." This off-book portfolio was of very low quality, consisting of related party lending and inadequate due diligence: some 82 percent of all outstanding loans did not meet the minimum required approval procedures, the necessary collateral, or appropriate documentation indicating a commitment or intention to repay obligations. Registered losses accumulated from 1989 to December 2002 in the Interbanco amounted to US$1.2 billion, and by March 2003, Baninter's overdraft amounted to US$2.1 billion (Procuraduría General de la República Dominicana, 2006). In effect, this double accounting system hid overdrafts, representing not only a failure in banking supervision but also fraudulent practices by the bank's managers. The Central Bank provided substantial liquidity support during the 2002-2003 period, leading to a severe decapitalization that was remedied through issuing Central Bank debt at high interest rates. Subsequent audits of the use of these bail-out funds to Baninter revealed further instances of fraud, such as increased lending to related parties and high-ranking bank employees, continued misreporting of liabilities, and repayment of related party deposits and obligations including those in offshore affiliates, all in violation of the Monetary and Financial Law (e.g., banks receiving liquidity support are prohibited from issuing new debt). The total direct cost of the bank bail-out to three failed commercial banks in 2003 amounted to 21 percent of GDP, and Baninter accounted for the lion's share--approximately 15-16 percent of GDP. The social and economic costs went beyond the direct cost of the bailouts: a rapid depreciation of the peso (due to a loss of confidence) led to very high inflation and a serious erosion of real incomes. Approximately 1.5 million Dominicans (about 16 percent of the population) fell below the poverty line in the aftermath of the banking crisis, 670,000 of whom fell into extreme poverty (World Bank and IDB, 2006). Although Dominican GDP growth has rebounded since 2004, the government and Central Bank continue to struggle to service the resulting higher debt stock. As of the end of December 2006, while the legal process continues, no one has been convicted of any crime related to the Baninter fraud. 25 3. RISK FACTORS FOR CRIME AND VIOLENCE Potential risk factors for crime victimization encompass conditions at the individual, relationship, community, and societal levels. As a whole Caribbean countries exhibit crime patterns similar to those in other countries. Both murder and robbery rates are higher in countries with low economic growth. Murder rates are also higher in poor countries and in communities that are poor and have large populations of young men. In Caribbean countries overall, homicide rates are 34 percent higher and robbery rates are 26 percent higher than in countries with comparable macroeconomic conditions. In general, the analysis implies that policies to reduce crime should focus on improving economic conditions, providing opportunities to young men, and improving trust in law enforcement. 3.1 This chapter assesses the risk factors for crime and violence in the Caribbean. Some studies characterize this type of analysis as one of determinants or correlates. The term "risk factors," which comes from the public health literature, is used here to emphasize that while it may be possible to cautiously infer causal relationships between crime and other variables, such relationships are not deterministic--the presence of a single risk factor does not guarantee that violence or crime will occur. In addition, the use of the risk factors terminology emphasizes that well-designed interventions may succeed in counteracting those risk factors. 3.2 An analysis of risk factors is potentially useful in a number of ways. First, a profile of the primary risk factors offers an understanding of who is at risk for criminal victimization. This may be useful in targeting interventions. Second, a profile of risk factors, when combined with theories of crime, can lend insights into the social process behind criminal activity, allowing for a better understanding of the nature of crime and ultimately leading to more effective anti-crime policy. CONCEPTUAL FOUNDATIONS OF RISK FACTORS FOR CRIME AND VIOLENCE 3.3 Economists have traditionally conceptualized criminal behavior as a rational decision made taking into account the expected benefits compared to the expected costs.46 The economic approach tends to lead to a focus on factors which directly affect the costs and benefits of participating in criminal activity: the relative returns of crime versus legal activities, the probability of being apprehended and convicted, and punishments for those convicted. 3.4 The analysis in this chapter seeks a wider perspective. A useful approach to conceptualize the various influences on crime and violence is the so-called ecological model (see Figure 3.1), which is broad enough to encompass a wide variety of theories from economics, sociology, and public health. The ecological model identifies four levels of influence on criminal and violent behavior. Individual factors include characteristics such as education level, marital status, and biological endowments. Relationship factors cover relations with peers, partners, and family. At a higher level, community factors 46A long line of research in this framework begins with Becker (1968). 26 include the broader context of social relationships in environments such as schools and neighborhoods. Finally, at the broadest level, there are societal factors such as cultural norms and economic conditions that also influence violence (WHO, 2002). Figure 3.1: Ecological Model for Understanding Crime and Violence Source: World Health Organization (2002). 3.5 This chapter presents two types of analysis. First, it shows results from a macro- analysis of risk factors for crime, using country-level data from around the world. Second, it provides insights from an analysis of household-level crime data in three Caribbean countries: Jamaica, the Dominican Republic, and Haiti. MACRO ANALYSIS OF RISK FACTORS FOR CRIME IN THE WORLD AND THE CARIBBEAN 3.6 Macro-level analyses can reveal how risk factors are associated with crime at the national level. These types of studies fall into two categories. Some studies examine trends over time in national crime rates in a single country and note how they are associated with changes in country-level variables. Others look at variation in crime rates across countries to determine how various factors may influence crime, sometimes drawing on panel data with multiple observations over time for particular countries. 3.7 The different approaches have various advantages and disadvantages. While a study of time trends in a single country can be tailored to take into account the particular circumstances of that country, in most circumstances it is difficult to infer a causal relationship between other variables and crime with national data from just one country. Studies across countries can exploit the large variation in crime rates and other conditions across the world. Cross-country studies suffer from several weaknesses, however. Cross- country crime data is often of questionable comparability; differences in official crime rates may reflect differences in definitions and reporting. Additionally, at best cross- country studies produce estimates of average relationships, which may be of limited relevance to individual countries. In principle, cross-country studies based on panel data should be of greatest interest. Only a few countries, however, produce long time series of crime data. 3.8 A few single-country studies have been conducted in the Caribbean. Ellis (1991) examines annual crime rates in Jamaica for the period 1950-84 and finds that high crime in the country is associated with periods of low economic growth, high unemployment, 27 and a large cohort in the 14-24 age group. He concludes that increases in crime in Jamaica can largely be explained as the consequence of economic decline, aggravated by changes in the population's age structure. Albuquerque and McElroy (1999) look at the case of Barbados and conclude that increased levels of property and violent crime there have been associated with worsening economic conditions and increased levels of visitor density. 3.9 This section presents a series of graphical bivariate comparisons of crime rates across countries with various risk factors and then proceeds to multivariate regression analysis. All the analysis is based on crime data collected from national police authorities by the United Nations Office of Drugs and Crime (UNODC) through its Crime Trends Surveys. The UNODC dataset used for the regression analysis includes three Caribbean countries--Barbados, Jamaica, and Trinidad and Tobago--along with 58 other countries. For the bivariate comparisons, the UNODC data is supplemented by additional data from other sources for some Caribbean countries (see data source description in Annex 3.1). 3.10 The analysis here is carried out using both homicide and robbery data. Many cross-country studies use only homicide data, because differences in reporting rates and definitions are thought to be minimal for this crime. Robbery is included as well to add a perspective on how risk factors for violent crime and property crime may differ.47 3.11 Figure 3.2 shows a collection of scatter plots for murder and robbery rates against various variables. For all the plots, each point corresponds to a country, and points corresponding to Caribbean countries are labeled. The number of points in each plot varies because some variables are only available for a selection of countries. In particular, the number of Caribbean countries for which robbery data is available is limited. 3.12 One pattern that is clear from the scatter plots is that both homicide and robbery rates exhibit strong inertia over time. In other words, countries that experience high crime rates in one period are very likely to have high crime rates in the following period. This can be seen in panels (a) and (b) of Figure 3.2, which show homicide rates in the period 1995-99 compared to those during the period 1990-94. There are several reasons why crime may show such persistence. First, criminal activity may become more attractive to individual criminals both as they become more experienced and as their legal labor market opportunities diminish with the accumulation of criminal records. Second, a society may fall into a high crime trap if, with a rise in criminal activity, increased social interactions with criminals reduce the social cost of criminal behavior. Likewise, if the legal system fails to respond to a spike in the incidence of criminal behavior, so that the perceived likelihood of arrest and conviction decreases, future criminal activity may rise as well. The inertia of crime rates also may reflect a link between crime and other variables which are persistent over time. Finally, the inertia pattern may reflect correlation over time in the errors in the data for individual countries. 48 47Robbery involves violence as well but can be considered a property crime with a violence component. 48For discussion of some of the reasons for crime inertia, see Glaeser, Sacerdote, and Scheinkman (1996), Leung (1995), Sah (1991), Posada (1994), and Mocan, Billups, and Overland (2005). 28 Figure 3.2: Cross-Country Correlates of Crime CRIME IS PERSISTENT OVER TIME a) Murder Rate: Late 90s vs. Early 90s b) Robbery Rate Late 90s vs. Early 90s 0 10 99 5-991 Jamaica Barbados n,oital 99-5991,noitla 0 Jamaica 10 puop pu 10 St.Kitts Barbados po 0 Trinidad & Tobago 0 00,0 00 0, 10 10 erpsedici R2=0.656 moH 1 10repsierebboR 1 R2=0.776 1 10 100 1 10 100 Homicides per 100,000 population, 1990-94 Robberies per 100,000 population, 1990-94 COUNTRIES WITH HIGHER GROWTH RATES HAVE LOWER CRIME RATES c) Murder Rate vs. Growth Rate d) Robbery Rate vs. Growth Rate 0 10 R2=0.055 ontilaup Jamaica noitla pu 0 Barbados Jamaica po po 10 000 Dominica Dominican Republic 0 Guyana 0,01reps Dominica 00,0 10 St.Kitts Barbados Trinidad & Tobago 10 Grenada Antigua & Barbuda deici erpseir 10 m be Ho obR R2=0.034 1 1 -5 0 5 10 -5 0 5 10 % Growth rate of GDP per capita % Growth rate of GDP per capita WEALTHIER COUNTRIES HAVE FEWER MURDERS AND MORE ROBBERIES e) Murder Rate vs. GDP per Capita f) Robbery Rate vs. GDP per Capita 0 10 R2=0.256 R2=0.135 ontial Haiti Jamaica noitla puop pu 0 Barbados Jamaica po 10 0 Dominica Dominican Republic 0 00,0 Guyana 00 Dominica 0, 10 10 St.Kitts erps Trinidad & Tobago Barbados Grenada Antigua & Barbuda deici m Ho 10repsierebboR 10 1 1 1000 5000 10000 30000 1000 5000 10000 30000 GDP per capita GDP per capita 29 (continued) Figure 3.2: Cross-Country Correlates of Crime COUNTRIES WITH HIGHER INEQUALITY HAVE HIGHER MURDER AND ROBBERY RATES g) Murder Rate vs. Gini Coefficient h) Robbery Rate vs. Gini Coefficient 0 10 ontilaup Jamaica Haiti noitla pu 0 Jamaica po 000 Dominican Republic po 10 0 Guyana 0,01reps 00,0 10 Trinidad & Tobago 10 deici erpseir 10 m be Ho obR R2=0.249 R2=0.041 1 1 20 30 40 50 60 20 30 40 50 60 Gini coefficient of income inequality Gini coefficient of income inequality COUNTRIES WITH MORE YOUNG MEN HAVE HIGHER MURDER RATES i) Murder Rate vs. Young Men in Population j) Robbery Rate vs. Young Men in Population 0 10 R2=0.087 R2=0.002 ontilaup Jamaica noitla Barbados pu 0 Jamaica po 000 Dominican Republic po 10 0 Guyana 0,01reps 00,0 10 Trinidad & Tobago 10 Barbados deici erpseir 10 m be Ho obR 1 1 12 14 16 18 20 22 12 14 16 18 20 22 % of population male age 15-34 % of population male age 15-34 MORE EDUCATED COUNTRIES HAVE LOWER MURDER RATES k) Murder Rate vs. Average Adult Education l) Robbery Rate vs. Average Adult Education 0 10 R2=0.289 R2=0.002 ontilaup Jamaica ontilaup 0 Jamaica Barbados po 000 Dominican Republic po 10 Guyana 000 0,01reps 10 Trinidad & Tobago Barbados 0,01reps deici deici 10 m m Ho Ho 1 1 2 4 6 8 10 12 2 4 6 8 10 12 Average years of education for adults Average years of education for adults Sources: Own analysis. UNODC data and other (see Annex 3.1 for discussion of homicide data source), Penn World Figures for growth rate in GDP per capita, Penn World Tables for GDP per capita in 1996 $US, Barro and Lee (1996) for education; World Development Indicators for population data by age group. 30 3.13 Countries with higher growth rates have on average lower rates of both homicides and burglaries (see panels (c) and (d) of Figure 3.2). This may indicate that higher growth reduces crime, by increasing opportunities in the legal labor market for those who might otherwise choose to engage in criminal behavior. The Caribbean countries in the sample have crime rates that are in all cases above the line showing the international tendency by GDP growth rate. In other words, on average countries in the Caribbean have higher crime rates than countries with comparable growth rates. 3.14 Countries with higher average incomes tend to have lower murder rates and higher robbery rates (see panels (e) and (f) of Figure 3.2). As with the growth rate plots, these figures show that each of the Caribbean countries for which data is available has a murder or robbery rate above the line showing the international tendency by GDP per capita. Higher inequality, on the other hand, is correlated with both higher murder and burglary rates. 3.15 The remaining panels of Figure 3.2 show overall correlations between homicide and two other factors that are often linked to high level of violence. Higher murder rates are found in countries with larger populations of young men and those with lower average adult education levels. Robbery rates, however, show no correlation with education levels or youth populations. 3.16 To attempt to isolate which variables may be driving crime rates, multivariate regression analysis is applied to the data. This work draws upon similar analysis conducted by Fajnzylber, Lederman, and Loayza (2000) (referred to as FLL in subsequent references) for the 1970-1995 period. Like FLL, this analysis is a based on an unbalanced panel dataset from all countries of the world with available data. Each observation is the average of values over a five-year period for a particular country. 3.17 The regression analysis presented here differs from the FLL analysis in several ways. First, it incorporates the updated United Nations crime data from the period 1995- 99. (The second half of the 1990s is the last five-year period for which the U.N. data is available for a large number of countries.) Second, it uses the simpler OLS regression approach, rather than more elaborate regression techniques.49 Table 3.1 presents results from the basic country-analysis. The explanatory variables used are the lagged value of the dependent variable, the growth rate of GDP per capita, average income per capita (in logs), and the Gini coefficient of income inequality. 49FLL employ the Arrellano and Bond generalized method of moments (GMM) estimator which attempts to control for endogeneity of crime determinants by instrumenting the explanatory variables with lagged values of the explanatory variables. 31 Table 3.1: Cross-Country Analysis - Basic Regression Results (1) (2) (3) (4) Homicide Robbery Homicide Robbery Variable rate rate rate rate Constant 0.914 ** -1.107 ** 0.998 ** -1.108 ** (0.458) (0.531) (0.453) (0.527) Lagged dependent variable 0.833 ** 0.902 ** 0.808 ** 0.891 ** (0.038) (0.031) (0.039) (0.032) Growth rate -0.017 * -0.026 ** -0.017 * -0.026 ** (0.010) (0.012) (0.010) (0.011) Average income (log) -0.089 ** 0.174 ** -0.094 ** 0.178 ** (0.040) (0.050) (0.039) (0.050) Income inequality 0.003 0.003 0.002 0.002 (0.005) (0.005) (0.005) (0.005) Caribbean dummy 0.343 ** 0.264 * (0.139) (0.148) Number of countries 61 54 61 54 Number of observations 182 179 182 179 R-squared 0.746 0.878 0.754 0.880 Source: Own analysis. UNODC for homicide and robbery rates; Penn World Figures for growth rate in GDP per capita and average income (per capita GDP); Deininger and Squire (1996) and World Development Indicators 2005 for Gini coefficient of income inequality. Notes: All regressions are OLS. 3.18 Like the bivariate comparisons, the regression results show that there is a strong element of inertia to crime rates and that higher crime rates are associated with lower growth rates. The regressions also show the same result as the scatter plot for average income: wealthier countries have lower homicide rates and higher robbery rates. In particular, the results from the simplest specifications show that a 10 percent increase in average income is associated with a decline of 0.89 percent in the murder rate and a 1.74 percent increase in the robbery rate. The effect of inequality on crime implied by the regression results, however, is not statistically significant after controlling for past crime, growth, and GDP per capita. 3.19 Columns (3) and (4) of Table 3.1 display results using the basic specification plus a Caribbean dummy. Those results show that for Caribbean countries in the panel sample (Barbados, Jamaica, and Trinidad and Tobago), homicide rates have been 34 percent higher and robbery rates have been 26 percent higher than in countries with comparable income per capita, growth rates, inequality, and past crime rates. Table A3.2 (in the Appendix) shows results from specifications which also include the average level of adult education and the percentage of the population made up of young men (age 15- 34). Neither shows a statistically significant effect on crime rates. The correlation of higher crime rates with lower education levels and large number of young men, apparent in the scatter plots, does not hold after controlling for past crime rates, the growth rate, and average income. 32 3.20 Overall, the cross-country analysis can be summarized as follows: ·There are strong inertia effects to crime, which means that once crime rates are high, it may be difficult to reduce them. At the same time, crime-reduction efforts in the short-term are likely to have huge long-term gains. ·As countries develop, violent crime tends to decrease, while property crime increases. ·Inequality is associated with both violent crime and property crime, but the relationship does not hold after controlling for other variables. ·Caribbean countries show similar patterns to the world as a whole, but have higher crime rates overall. 3.21 Why do Caribbean countries have higher crime rates on average? The obvious candidate reason is drug trafficking. Although the countries of the Caribbean are very diverse along many lines, one thing they have in common is that nearly all are used as points for drug transshipment. De Albuquerque and McElroy (1999), among others, have noted that the rise of crime in the Caribbean over time coincides with the expansion of the narcotics trade. The U.S. Government's most recent annual report on the international drug trade, Department of State (2006), lists nearly every country in the Caribbean as a "transshipment point."50 Large quantities of drugs are regularly seized by law enforcement in most Caribbean countries; of the top 20 countries in the world in terms of drugs seizures per capita, half are in the Caribbean.51 The Dominican Republic, Haiti, Jamaica and the Bahamas are all identified by the U.S. report as "major" drug transit countries. While the drug trade clearly does not explain all variation in crime rates across countries, it is undoubtedly an important contributing factor to crime and violence in the region. 3.22 It is very important to recognize the limitations of regression analysis. First, regression analysis provides estimates of average relationships between variables. In the case of cross-country regressions, the relationships are averages across countries of the world. There are inevitably exceptions which do not fit the average pattern. To take just one example, while the regression results show that higher growth rates are associated with lower crime rates, some countries experience both rapid growth and high crime rates, as has been the case for Trinidad and Tobago in recent years. But the regressions do control for other factors, which helps clarify the relationship between crime and growth. 3.23 It should also be noted that with all the relationships presented here there are multiple possible relationships between the variables. It may be the case that causality runs not from growth rates to crime but in the other direction, so that a higher rate of violence tends to decrease growth. Viewed from this perspective, the growth rates of 50Specifically, the report lists the following countries as transshipment points: Antigua and Barbuda, Aruba, the Bahamas, Barbados, Cuba, Dominica, the Dominican Republic, Grenada, Guyana, Haiti, Jamaica, the Netherlands Antilles, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, and Trinidad and Tobago. Given that such a large number of Caribbean countries are classified as transhipment points, it is not possible to introduce a "drug transhipment" dummy variable in the cross-country regressions; such a dummy would be highly collinear with Caribbean region dummy. 51This statement is based on analysis of the UNODC's drug seizure database. 33 Caribbean countries are higher than would be predicted based on their crime rates alone. It is also possible that the negative correlation between growth rates and crime rates is due to a third factor which affects both growth and crime. 3.24 Econometrics offers a variety of strategies for attempting to distinguish causal relationships from mere correlations. Unfortunately, convincing applications face a number of hurdles, foremost among them being data requirements that are difficult to fulfill for cross-country crime regressions. The analysis in this chapter does not apply such techniques but instead uses the simplest form of regression analysis, ordinary least squares. Rather than making hard claims of causality, the analysis in this chapter is presented in the spirit of providing suggestive evidence on possible relationships between crime and other variables. Note that the primary finding from the cross-country analysis--the fact that the crime rates of Caribbean countries are above the rates predicted by purely economic variables--does not rely on any causal story, but rather it suggests that given its economic performance, the Caribbean's crime rates are on average higher than what would be expected. 3.25 In Chapter 4 of this report, cross-country analysis is presented on the possible effects of growth on crime. In that case, the focus is necessarily on identifying a causal relationship, but the obstacles to doing so are no less daunting.52 Also, the identification of any "causal" effect of crime on growth depends on the assumption that past levels of crime are significantly different from current levels of crime, and in any case, this technique relies on technical statistical properties of the data rather than on an intuitive or conceptual identification of the causal effect of crime on growth. Consequently, readers should be modest in drawing strong causal conclusions from this analysis as well. RISK FACTORS FOR CRIME AT THE HOUSEHOLD LEVEL: EVIDENCE FROM THREE CARIBBEAN COUNTRIES 3.26 While macro-level estimates are useful for providing a sense of average relationships between crime and other variables, country-specific studies based on microdata are useful for examining the circumstances in individual countries. Micro-level studies also allow for a more detailed examination of risk factors at the community, relationship, and individual level. This section considers risk factors for crime victimization, using victimization data collected at the household level in three countries: Jamaica, the Dominican Republic, and Haiti. The victimization data for Jamaica comes from the national population census conducted in 2001. The Dominican Republic data is from a household survey in 2005, matched with information from the national census at the province level. Finally, the Haiti analysis is based on more limited crime victimization questions from a 2001 survey. 3.27 For all three countries, the micro-level risk factor analysis uses probit regressions, with crime victimization as an outcome variable and household characteristics as explanatory variables. The set of explanatory variables varies by 52The growth-on-crime analysis employs the Arellano-Bond estimator that is sometimes used, particularly in the cross-country literature, to identify a causal relationship. When measurement error in the data is likely to be substantial, as is typically the case for crime data, the estimator is less reliable. 34 country but includes both household and community-level characteristics. Complete regression results from each country can be found in Annexes 3.2, 3.3, and 3.4. Results from the regression analysis are summarized in Table 3.2 and described below. Table 3.2: Summary of Micro-Analysis of Risk Factors for Criminal Victimization Dominican Jamaica Republic Haiti Property Property Violent Property Crime Crime Crime Crime (Burglary/ (Burglary/ (Murder) (Robbery) Mugging) Mugging) Household-level Consumption/income - + + + Female-headed + - not sig. not sig. Young men + - not sig. not sig. Education level + - + not sig. Community/province-level Consumption/income - + not sig. n/a Inequality not sig. not sig. - n/a Female-headed + + - n/a Young men + + + n/a Education level - - + n/a Population density + - + n/a Urban + + + + Reporting rate - - not sig. n/a Source: Own analysis. Notes: Table entries indicate results from multivariate regression results for each variable, for analyses from different countries and crimes. A plus sign indicates a positive relationship, a negative sign indicates a negative relationship, "not sig." indicates that results for the variable in question were not significant at the 10 percent level, and "n/a" indicates that the variable was not available for a particular country. See Annex 3.2, 3.3 and 3.4 for detailed results. For Jamaica, results shown for violent crime are for murder and those shown for property crime are for robbery. Economic Factors 3.28 For Jamaica, the patterns of crime victimization by consumption level--of both households and communities--mirror the trends in the cross-national data: increases in wealth are associated with lower levels of violent crime and higher levels of property crime. In Jamaica, wealthier household are much more likely to experience property crime (robbery and the theft of agricultural goods), while poorer households are more at risk to be victims of all violent crime. Additionally, households in poorer areas, even after controlling for the household-level consumption, suffer higher risk of murder and wounding. On the other hand, households in wealthier areas are more likely to be victims of property crime. This last relationship can be seen in Figure 3.3, which shows a scatter plot of robbery rates versus mean consumption by neighborhood (census enumeration district.) 35 3.29 In the Dominican Republic, better off households are more likely to suffer from all forms of theft. Even after controlling for household-level welfare, all types of theft occur more frequently in wealthier provinces. Provincial-level unemployment is also associated with higher levels of personal theft. Figure 3.3: Robbery Rates vs. Mean Consumption by Neighborhood Property Crime is More Prevalent in Wealthier Areas in Jamaica Source: Analysis of 2001 Population and Housing Census. Notes: For the purposes of the figure, observations with values of zero for robberies were replaced with the smallest value in the sample. 3.30 In Haiti, wealthier households are more likely to be victims of both burglary and mugging. The distribution of victimization by quintile is shown in Figure 3.4. Robberies, burglaries, and muggings disproportionately strike Haitians in the top two quintiles, who are most likely to have property of value. Personal injury victims, however, are concentrated in both the top and bottom quintiles. 3.31 The analysis does not find a consistent relationship between crime and inequality. In the multivariate regressions, all but one of the crimes show no relationship to community inequality levels; only the theft of agricultural products from farms is found to be more prevalent in higher inequality areas. Counter to expectations, in the Dominican Republic vehicle theft and burglary occur less often in provinces with higher inequality. These findings differ from those from a similar study in South Africa (Demombynes and Özler, 2005) which found local inequality to be correlated with both violent and property crime, after controlling for plausible proxies for the returns from crime and non-crime activities. 36 Figure 3.4: Distribution of Victimization by Crime and Quintile in Haiti 0% 20% 40% 60% 80% 100% Robbery 13% 18% 18% 28% 23% Burglary 14% 16% 18% 23% 30% Injured 35% 10% 14% 13% 29% Mugging 10% 17% 18% 25% 30% Poorest Q_2 Q_3 Q_4 Richest Source: Own analysis of ECVH 2001. Note: The figure shows the percentage of victims of each crime who are in each income quintile. Demographic Factors 3.32 In Jamaica, households in communities in which female-headed households make up a larger share of the population are more likely to suffer from murder, shooting, and robbery. This finding is compatible with sociological theories that predict that female-headed households are more prevalent in areas with breakdowns of social ties, which may lead to more violent crime. The opposite relationship holds, however, for mugging in the Dominican Republic, and across all three countries there is no consistent relationship at the household level between victimization and female headship. 3.33 The presence of young males in Jamaican households is associated with higher rates of violent crime victimization, and the presence of more young males in the district is associated with higher homicide rates. More males in the household, however, are also associated with lower rates of property crime, which suggests that they may serve a protection function for the household, even though they are highly likely to be victims themselves. 3.34 In the Dominican Republic, the presence of young males in the household is not significantly associated with crime victimization, but the presence of more young men in the province is strongly associated with all forms of property crime captured in the survey. In Haiti, however, there is no clear relationship between age structure and the likelihood of criminal victimization. 37 Box 3.1: Haiti's Entrepreneurs of Violence Political violence is not a novelty in Haiti's history, and it neither started nor ended with the Duvalier regime, although this regime's violent record was unprecedented. "Papa Doc" Duvalier (1957-71) began to institutionalize political violence soon after he was elected in 1957, by establishing a force of cagoulards ("hooded men") charged with silencing supporters of rival candidates (who challenged the election results), as well as other dissidents, which gradually developed into a more extensive network of spies for the Duvalier regime. More recently, Haiti's democratic experience has failed to provide capable and stable governance. Initially, democracy was brutally repressed by the army and their supporters, using armed paramilitary groups that have been a feature of Haitian politics since Duvalier's macoutes. From Aristide's return in 1994, democracy was undermined by deep conflicts among erstwhile democratic allies, resulting in a political stalemate that lasted almost from Aristide's return in 1994 until his departure in 2004, and which undermined economic growth and state building alike. Moreover, it transformed important parts of the democratic movement, the urban popular organizations, into agents for the Lavalas party, using brutal violence in struggles over territorial control and state favors. With Aristide's departure, some of these groups are bent on undermining the political process, in collusion with some former political leaders. Their area of operation and influence is in the slums and "popular areas" that provide a rich reservoir of potential recruits. Violence and insecurity in these areas is pervasive and these "entrepreneurs of violence" are capable and willing to engineer violence for political or other purposes, as the line between political and criminal violence has become increasingly blurred. Source: "Urban Violence in Haiti: A rapid assessment of socio-economic, political, and institutional causes of urban violence in Haiti." (World Bank, 2005) Education 3.35 Jamaican households in less educated areas are more likely to be victims of all crimes. This may reflect the fact that the less educated have worse legitimate wage- earning opportunities and hence the supply of potential criminals in such areas is larger. Patterns are less consistent for the education level of the household itself. Households with more educated adults suffer slightly more shootings and woundings but fewer robberies relative to the less educated. 3.36 The Dominican Republic analysis shows somewhat different results. Households living in more educated provinces are at greater risk of muggings, while more educated households are slightly more likely to be victims of burglary and mugging. The Haiti results show no pattern by education level of household members. Urban vs. Rural and Population Density 3.37 In all three countries, crime is chiefly an urban phenomenon. In both Jamaica and the Dominican Republic, rates of victimization are highest in urban areas for nearly all crimes, (an exception is crop and livestock theft in Jamaica, which is necessarily more likely to occur in rural areas). These results hold even after controlling for population density. In Haiti, robbery, burglary, and mugging rates are highest in urban areas. All the crimes captured in the victimization survey in the Dominican Republic are more 38 prevalent in areas with higher population density. In Jamaica, however, population density is positively associated with violent crime but negatively associated with property crime. Crime Reporting Prevalence 3.38 An important finding of this study is that in Jamaica, the reporting prevalence is negatively associated with crime rates. In other words, crime rates (as measured by victimization data) are higher where a lower percentage of crimes are reported to the police. This has multiple implications. First, this suggests that official police data distort the true geographic profile of crime, because official data are biased downwards for higher crime areas. Second, the reporting rate can reasonably be plausibly be interpreted as a measure of confidence in the police, as people will be more likely to report when they trust the police and believe they will respond. If this interpretation is correct, lack of trust and confidence in the police is lower in areas with higher local crime rates. 3.39 This correlation could be explained by a number of different causal relationships. One possibility is that lower trust in the police leads to higher crime, because the police are less effective when they cannot count on the support and assistance of the population they serve. But causality could well be in the reverse direction, with higher crime leading to lower levels of trust in police. This could be the case if people form their levels of trust based on their observations of local crime levels. Finally, both distrust and high crime rates could be correlated with a third factor not controlled for in the regression analysis. An unobserved variable, like a high level of social exclusion, could both drive high crime rates and inspire lower confidence in the police.53 Summary of Micro-Analysis 3.40 Overall, several key points emerge. Although the precise relationship varies by country, local economic conditions are associated with crime. In particular, wealthier areas are consistently more likely to be the targets of property crime. In Jamaica, where information on both violent crime victimization and local conditions are available, violent crime strikes poor more than rich neighborhoods. Local inequality and crime victimization do not, however, have a clear relationship across countries. 3.41 The results related to demographic characteristics are mixed. Everywhere, crime is concentrated in urban areas, and crime rates are higher where there is a larger population of young males in the Dominican Republic and Jamaica (for violent crime). Areas with large number of female-headed households in Jamaica suffer higher crime rates, but that relationship does not hold for the Dominican Republic. 3.42 In Jamaica, community levels of trust in the police, as proxied by reporting rates, are related to crime rates. This suggests that measures that increase the trust in police, or address the root causes for the lack of trust in the police, may help reduce crime. 53Soares (2004) examines crime reporting rates at the national level, finding that reporting rates are highest in countries with longer periods of institutional stability, more police per capita, and lower levels of corruption as measured by subjective indices. 39 3.43 It is important to note that the explanatory power of these regressions is fairly low.54 While many risk factors are strongly correlated with crime victimization, it is still largely a random process from the perspective of potential victims. Consequently, the costs of the threat of crime are universal. POLICY IMPLICATIONS 3.44 The analysis in this chapter points to a number of potential policy conclusions. First, the persistence of crime over time shown by the cross-country regressions suggests that efforts to reduce crime in the short run may have very large long-term effects. Second, while the direction of causality between economic conditions and crime is subject to debate, the weight of evidence suggests that increasing levels and growth rates of per capita income reduce violent crime. At the same time, rates of property crime are higher in both countries and communities with higher incomes. 3.45 Additionally, the presence of large populations of young men is associated with higher crime rates in communities. This suggests strongly that crime prevention interventions should be targeted at young men. 3.46 Finally, the fact that crime rates are higher in areas in Jamaica where trust in the police is lower suggests that policies that improve trust in the police, or address the underlying causes of a lack of trust in the police, can help reduce crime. Policies which improve the transparency of the police, reduce corruption, and make the police more accountable and accessible to citizens may be effective routes to bringing down criminal activity. 54The highest pseudo R-squared values are 0.077 in Jamaica (for crop and livestock theft), 0.084 in the Dominican Republic (for mugging), and 0.068 in Haiti (for mugging). R2 values refer to the percentage of variation in the dependent variable which can be explained by the independent variables in the regression. In other words, the higher the value of the R2 , the better job the explanatory variables do in explaining the dependent variable. 40 4. SOCIOECONOMIC COSTS OF CRIME There are a wide variety of approaches that have been used to catalog and measure the costs of crime and violence. Analyses for the Caribbean have found that fear of crime causes individuals in Jamaica, the Dominican Republic, and Haiti to avoid activities and locations that are perceived as exposing them to risk of being victimized by crime. People whose families have suffered from crime report substantially lower levels of life satisfaction, and this effect is much greater in the Dominican Republic than in non- Caribbean countries. Crime also reduces tourist arrivals in the region, discourages business investment, and stifles economic growth. Cross-country panel data suggests that Haiti and Jamaica could boost economic growth per capita by 5.4 percent per year if they were to bring their homicide rates down to the levels of Costa Rica. Guyana and the Dominican Republic would also benefit substantially, with potential growth rate increases of 1.7 percent and 1.8 percent respectively. 4.1 The literature on crime and violence covers a wide ground, and there are multiple typologies for the possible sorts of costs that could be contemplated. The World Health Organization (2004) draws a distinction between direct costs (medical, legal, policing, prisons, foster care and private security) and indirect costs (lost earnings and time, lower human capital, lower productivity, lower investment, psychological costs and other non-monetary costs). Buvinic and Morrison (1999) use a more complex typology and distinguish between the following: ·Direct costs: the value of all goods and services used to prevent violence or offer treatment to its victims or perpetrators. This has been the most commonly estimated category of costs and includes health costs, police, justice and prison costs, as well as resources spent on private security measures. While the most frequently measured, this category may not be the most important. ·Non-monetary costs: higher mortality and morbidity rates that result in pain, suffering and death, but not necessarily result in either expenditures on health care or in easily quantifiable economic losses. ·Economic multiplier effects: impacts on human capital, labor force participation, lower wages and incomes, savings and macroeconomic growth. ·Social multiplier effects: erosion of social capital, inter-generational transmission of violence and lower quality of life. 4.2 One also might distinguish between short-run costs of all sorts and the long-run effects on growth. An earlier report on youth in the Caribbean by the World Bank (2003a) lists the following costs: 1) arrest, prosecution, and detention of criminals; 2) property loss and damage; 3) medical costs, public programs for victims, and lost income of the victim; 4) intangible costs (pain, suffering, and quality of life); 5) security costs; 6) lower tourist receipts; 7) lost income due to incarceration; and 8) lost social capital. 41 APPROACHES TO MEASURING THE COSTS OF CRIME AND VIOLENCE 4.3 There are a number of different methodologies that have been employed to assess the costs of crime and violence. These include: 1) assessing specific costs; 2) adding up total costs using an "accounting approach"; 3) estimating total cost (more accurately, willingness to pay) using econometric methodologies; 4) calculating the Disability-Adjusted Life Years (DALYs) lost due to violence; and 5) estimating effects on economic growth through cross-country regressions.55 4.4 Each of the methodologies has its relative strengths and weaknesses. Studies which look at specific costs analyze the effects of crime on particular sectors or population groups without attempting to be comprehensive--i.e., without attempting to calculate the total cost of crime to society. 4.5 Among economists, the more popular approach has been to summarize the various costs of crime and violence in a single monetary figure that purports to represent the total cost associated with crime for a particular country. This is frequently done using an accounting approach, which collects data on the costs of crime in disaggregated categories and then sums up over categories (Buvinic and Morrison, 1999). This generally involves adding up all the assorted public and private expenditures associated with crime. 4.6 The accounting approach has both advantages and disadvantages. On the one hand, a single number may be useful for summarizing the myriad costs, for providing a comparison point to assess the cost-effectiveness of interventions, and for communicating to economists and business interests used to thinking in monetary terms. It is also attractive in data-poor environments; if information is missing for some categories of costs, estimates can be generated using categories for which data are available (Buvinic and Morrison, 1999b; World Bank, 2006b). On the other hand, no single measure can capture all imaginable costs and any specification of categories is necessarily arbitrary. 4.7 Another approach--which does attempt to measure the marginal benefit of crime reduction--is to estimate individual willingness-to-pay for marginal reductions in crime rates. This has been done in the case of Brazil, for example, by using hedonic housing models to estimate a relationship between housing prices and crime rates (Hermann and Haddad, 2003). It also could be done using contingent valuation methodology, which uses surveys to measure individuals' willingness to pay for reductions in crime. 4.8 A fourth approach is to estimate disability-adjusted life years (DALYs) lost to violence. The World Health Organization (2002) has estimates of violence-related DALYs lost for a large number of countries; DALYs are a widely-accepted measure of the health costs of violence. 55There are also studies which attempt to measure the marginal willingness-to-pay for reductions in crime. No such study has been conducted in the Caribbean. 42 4.9 A fifth method is to estimate the effect of crime on growth using dynamic panel regressions. The growth effects approach has the advantage that is a summary measure which in principle captures the ultimate long-run effects of crime through many channels. 4.10 Rather than choose a single methodology to measure the costs of crime and violence, this chapter reviews a variety of work on its costs in the Caribbean, including several studies that estimate specific costs of crime, an accounting approach to measuring total costs, and estimates of DALYs lost due to violence. Also presented are new estimates of the effects of crime victimization on a self-reported measure of life satisfaction. Finally, the chapter builds on recent (2006) World Bank estimates of the impact of violent crime on growth to calculate the "growth dividend" that could be produced by reducing the homicide rate in several Caribbean countries. SPECIFIC COSTS OF CRIME 4.11 Several studies in the Caribbean examine specific costs of crime. These necessarily cover a wide variety of types of costs. Considered here are those that address the effects of crime on tourism, the non-monetary effects of violence in poor urban neighborhoods in Jamaica, gender-based violence in Haiti, and private sector responses to crime. Tourism 4.12 Because of the key role that tourism plays in many Caribbean countries, the effects of crime on tourism are of particular concern. In 2004, the Jamaican tourism minister said that Jamaica's unprecedented crime level was threatening to derail the tourism industry by scaring away visitors and hurting investment (Associated Press, 2004). A study of popular perceptions of those working in the tourism industry found that crime and violence were perceived as the main problem afflicting the tourism industry (Dunn and Dunn, 2002). 4.13 A few studies consider the relationship between crime and tourism for the Caribbean. Albuquerque and Elroy (1999) showed that property crime tends to be disproportionately directed at tourists. Nonetheless, King (2003) showed that the odds of being victimized as a tourist in the Caribbean are low. He suggested that perceptions of the danger of crime do affect tourism but that such perceptions are driven much more by mass media coverage in the sending country than by actual dangers. 4.14 Alleyne and Boxil (2003) examined the relationship over time between tourist arrivals and crime in Jamaica and concluded that crime has discouraged tourists, particularly from European countries, but that the negative effect of crime has been mitigated by increased advertising and promotion and the growth of all-inclusive hotels. All-inclusive hotels, however, are incapable of fostering significant backward linkages to the rest of the local economy. A new business model for tourism that goes beyond airlines, cruise lines and hotels to span the entire "destination experience"--restaurants, taxis, tour operators, cultural and heritage sites and scenic locations--is necessary to 43 generate backward linkages (World Bank, 2006a). This new model, however, cannot succeed if tourists are unwilling to leave all-inclusive resorts because of fear of crime. Non-Monetary Effects in Jamaica and the Dominican Republic 4.15 Other qualitative work has examined the non-monetary effects of violence. A participatory study in poor urban Jamaican neighborhoods (Moser and Holland 1997) considers the effects of violence. The study notes that residents of inner-city neighborhoods in Jamaica suffer from "area stigma." They are judged to be associated with criminals based on where they live, which makes it difficult for them to find employment. The study also considers the intangible effects of violence on social capital, noting that "violence erodes social relationships, not only through death, but by restricting physical mobility and increasing levels of tension." Due to the high levels of violence in parts of urban Jamaica, residents are afraid to leave their homes and interact less often with friends and family who live elsewhere. 4.16 A similar phenomenon is evident in responses to the Encuesta Nacional de Hogares de Propósitos Múltiples (ENHOGAR) survey conducted in the Dominican Republic in 2005. Surprisingly large numbers of respondents report that they have stopped doing many activities due to fear of crime. As can be seen in Figure 4.1, due to the high levels of delinquency, Dominicans have reduced their recreational activities outside the house. Figure 4.1: Responses to Fear of Crime in Dominican Republic -What Do People Stop Doing Due to Fear? Leave the house 19.4% Entertainment activities 12.8% Going out at night 12.2% Stop visiting friend or relatives 7.0% Participation in activities 2.5% Vacations 2.2% Stop w orking 2.1% Others 1.9% Stop studing 1.6% Doing exercise 0.8% Going to church 0.4% 0% 5% 10% 15% 20% 25% Source: Own analysis of ENHOGAR 2005. Note: Shown are the percentages responding in the affirmative to "Which of the following activities have you stopped due to fear to crime?" 4.17 Similarly, survey data from Haiti show that a high percentage of respondents are afraid of going to specific places due to fear of being mugged, attacked or having money or other belongings stolen. Figure 4.2 depicts that Haitians are more likely to avoid places like Port-au-Prince or the nearest larger town. 44 Figure 4.2: Responses to Fear of Crime in Haiti: Where Do People Stop Going Due to Fear? To Port-au-Prince 41.9% To the nearest larger tow n 32.3% To the local market 27.8% To visit people in this tow n 27.2% To the bank 24.4% To the neighbors 22.3% To w ork 20.2% To a bar 19.3% To a rooster-fighting arena 16.3% To the lottery office 16.2% To a night club 10.8% 0% 10% 20% 30% 40% 50% Source: ECVH 2001. Note: Shown is the percentage responding in the affirmative to "Are you ever afraid of going to any one of the following places out of fear of being mugged, attacked or having money or things stolen from you?" Gender-Based Violence in Haiti 4.18 Morrison and Orlando (2005) examine the impact of gender-based violence in Haiti on women's health and employment and children's health, using propensity-score matching techniques. Results from this analysis for lifetime physical violence by an intimate partner are shown in Table 4.1.56 Key results are that suffering from physical violence is strongly associated with an increased likelihood that: i) women do not receive antenatal care; ii) women suffer from genital sores and ulcers; and iii) both women and their children suffer from anemia. 56Lifetime physical violence is physical violence suffered at any point in an individual's life. 45 Table 4.1: Effects of Lifetime Physical Violence by Intimate Partner in Haiti Average net effect of lifetime physical OUTCOME VARIABLE violence by intimate partner (% increase or decrease compared to non-victims) WOMEN'S HEALTH Weight for Height (centimeters x kilograms) -1,99% Anemia (severity degree 0-3; 0=no anemia; 3=severe anemia) 27,63%* Number of Children -5,46% % of Women who Ever had a Terminated Pregnancy 33,18% Last Child Wanted (index 1 =wanted ­3=did not want more children) 5,53% Sexually Transmitted Disease (%) 55,12% Genital Sore /Ulcers (%) 116,22%* WOMEN'S USE OF HEALTH SERVICES Visited Health Facility (%) 19,68% Antenatal Care (%) -17,88%** Births Assisted by Health Care Professional (%) 1,88% Unmet Family Planning Needs (%) 8,57% Contraceptive Use (%) 22,16% WOMEN'S EMPLOYMENT Employed and Earning Cash (probability) 2,27%* CHILDREN'S HEALTH (AGES 0-5) Diarrhea (%) 25,19% Coughing (%) 13,83% Anemia (%) 31,81%* Height for Age (centimeters x age in months) -5,79% Weight for Height (centimeters x kilograms) -3,41% Immunization (%) -13,73% Under 5 mortality (per 1000 births)# 17,43% CHILDREN'S EDUCATIONAL ACHIEVEMENT (AGES 7+) Education Years -12,21% Education Gap 7,95% School Attendance (%) -3,96% Source: Propensity score matching estimates from Morrison and Orlando (2005), based on analysis of Haiti Demographic and Health Survey Data. * significant at 10% ** significant at 5% *** significant at 1% #This is not a population child mortality rate. This variable is the sample average of the number of children under 5 who died divided by each woman's total number of births per 1000. In this case both rates are similar since there is no significant difference in the total number of children between the victims group and the control group. 46 Costs of Crime to Business 4.19 The effects of crime on businesses can be particularly damaging because they can involve both short-run costs and long-run consequences for development by diverting resources to crime prevention measures and otherwise discouraging investment. This section reviews key findings from surveys of businesses on the costs of crime in Jamaica and the Dominican Republic. The survey in Jamaica was carried out by Francis et al. (2003) as part of the background work for the 2003 World Bank Country Economic Memorandum (CEM). The results from the crime module of the 2005 Investment Climate Survey (ICS) in the Dominican Republic are presented in the country's most recent CEM, World Bank (2006a). Figure 4.3: Impact of Crime on Various Business Practices in Jamaica Increased cost of security 51% Negative impact on plans for business expansion 39% Negative impact on investments to improve 37% productivity Negative impact on worker productivity 22% Increased cost of services purchased 18% Other 13% 0 10 20 30 40 50 60 % of Firms Reporting "Significant", "Somewhat Significant", or "Highly Significant" Impact Source: 2001 Firm Victimization Survey, described in Francis et al. (2003). 4.20 Managers interviewed for the Jamaica study described how crime affected their business practices. Figure 4.3 shows the percentages of managers indicating that crime had either a significant, somewhat significant, or highly significant impact on particular business practices. Unsurprisingly, security clearly dominates the impact of crime, with more than half of firms reporting that crime increased security costs. Managers also indicated, however, that crime affects business decisions in ways that are likely to affect output in the long run. Thirty-nine percent responded that they were less likely to expand their business because of crime, and 37 percent reported that crime discourages investments that would improve productivity. 47 Figure 4.4: Crime Protection Measures Taken by Firms in Jamaica Installing grill 58% Upgrading accounting/records 50% Special fencing of premises 49% Hiring unarmed security guard(s) 49% Internal security system 37% Closing before dark 36% Installing electronic alarm system 31% Hiring armed security guard(s) 31% Hiring community protection 9% 0 10 20 30 40 50 60 70 % Using Measure Source: 2001 Firm Victimization Survey, described in Francis et al. (2003). 4.21 The steps taken by businesses to protect themselves from crime are varied and not limited to security measures narrowly defined. A summary of these responses is shown in Figure 4.4. Physical security measures are the most common: 58.3 percent have installed protective grills on buildings, 49.3 percent have special fencing, and 31 percent have installed alarm systems. Nearly half have unarmed guards, and a remarkable 30.8 percent of firms have armed guards. Many firms (36 percent) have opted to close before dark; this practice has especially high costs in the manufacturing sector, where second shifts are not used and productive capacity consequently sits idle. 4.22 Among the more troubling costs to business are extortion and protection costs. Extortion occurs when a firm pays an extortionist in order to avoid victimization threatened by the extortionist. A protection racket, though similar, is distinguished in that the payment is in exchange for protection from criminal behavior from all other sources. To capture both types of coerced payments, the survey inquires about both. Notably, 8.5 percent of managers in Jamaica list paying for protection as being among their crime prevention measures, and 5 percent reported that they were force to pay extortionists. 4.23 Data from the firm victimization survey shows that security costs pose the largest burden for small firms. As shown in Figure 4.5, small firms pay security costs equal to 17 percent of their revenue on average, as compared to just 2 percent of revenue for firms overall. 4.24 Findings regarding crime from the ICS in the Dominican Republic were broadly similar to those from the Jamaica study. Seventy-two percent of firms surveyed took at least one action to prevent crime. As in Jamaica, substantial numbers of firms reported that they installed security equipment or software, hired security personnel, and closed before dark in response to the threat of crime (see Figure 4.6). 48 Figure 4.5: Private Security Costs for Firms by Size of Enterprise as Percentage of Firm Revenue in Jamaica 18 17.0% eu 16 en evR 14 of % 12 as esrutdi 10 enpxEytiru 8 7.6% ecS 6 etavriP 4 2.0% 2 0.7% 0 All Firms Small Firms Medium- Large Firms Sized Firms Source: 2001 Firm Victimization Survey, described in Francis et al. (2003). Figure 4.6: Crime Protection Measures Taken by Firms in the Dominican Republic Install security equipment 40% Hire security personnel 39% Close before dark 32% Install security software 27% Other actions 10% 0 5 10 15 20 25 30 35 40 45 Percentage Using Measure Source: World Bank (2006a), based on 2005 Investment Climate Survey. 4.25 Firms in the Dominican Republic also report other impacts of crime. Sixty-three percent cite crime as a major obstacle to investment. Similarly, 57 percent indicated that their access to financing had declined as result of crime, although only 10 percent of the firms were actually victims of crime, illustrating the fact that the effects extend far beyond the direct effects on victims. Other impacts of increased crime mentioned by 49 respondents to the ICS include increases in spending on formal and informal security measures and declines in worker productivity (see Figure 4.7). Figure 4.7: Impact of Higher Crime on Businesses in the Dominican Republic Lower worker productivity 47% Higher spending on informal anti-crime 54% measures Higher spending on formal anti-crime measures 36% Worse access to financing 57% 0 10 20 30 40 50 60 % of Firms Reporting Moderate or Significant Impact Source: World Bank (2006a), based on 2005 Investment Climate Survey. TOTAL COSTS: ESTIMATES USING AN ACCOUNTING APPROACH 4.26 Estimates using the accounting approach for measuring costs vary in terms of the extent of costs they measure. As part as the background work for this report, Holder and Mutota (2006) estimated that the costs of crime in Trinidad and Tobago in 2003 amounted to TT$1.098 billion (US$160 million), 1.6 percent of 2003 GDP. This estimate includes the value of lost productive years due to both fatal and non-fatal injuries, the value of the lost productivity of ex-criminals who have reduced earnings capability after serving jail time, funeral costs, and business security costs. It does not include the costs of public security and other elements of the criminal justice system. 4.27 An attempt at a comprehensive study of the costs of crime in Jamaica was conducted by Francis et al. (2003). They included both private and public costs in their summary measure. The study considered health costs--both those borne by the public health system and those paid by private citizens--along with the value of lost production due to mortality and injury related to crime, public expenditures on security, and private expenditures on security. 4.28 Health costs are among the most important costs of crime and violence in Jamaica. Violence has reached such high levels that it puts a substantial strain on medical services. A study of hospital-based data suggested that violence is the chief source of injuries in Jamaica. Data collected at three hospitals in 1999 and 2000 showed that 51 50 percent of injuries were violence-related, with the remainder classified as unintentional (33 percent) and motor vehicle-related (15 percent) (Sharon Arscott Mills, 2002). Other studies in Jamaica show similar figures.57 4.29 The Francis et al. study estimated the public health costs of violence by calculating the share of emergency room admissions that are due to violence-related injuries and then multiplying that share by the total cost of the public health system. Using this method, they find that the total annual cost to the public health system is J$996 million or about 0.3 percent of GDP in 2001. 4.30 Private health costs for those hospitalized due to injury are estimated based on data from the 1998 Jamaica Survey of Living Conditions. Using data on the average cost of hospital stays, treatment, and medication for those hospitalized for violence-related injuries, along with the number hospitalized for such injuries, the total private costs from violence-related hospitalization are found to be J$254 million, just under 0.1 percent of GDP in 2001.58 This estimate does not include the private costs of treatment to those not hospitalized. 4.31 Apart from direct private health costs, some victims of violence are incapacitated and unable to carry out normal activities during periods of convalescence. Survey data suggest that the average length of hospital stay for violence victims was nine days. Using information on the average wage and number of people hospitalized for violence, and assuming that a hospitalized victim is out of the workforce for two weeks, the value of time lost due to violence-related morbidity is estimated to be J$337 million, about 0.1 percent of GDP.59 4.32 Another part of the cost of violence is the funeral costs for murder victims. Based on a poll of funeral service providers, the study estimates funeral costs to be J$66,000 per person average, for a total cost of J$64 million. The study also counts the time required for funeral preparations, which are found to have a value of approximately J$41 million. 4.33 Additionally, the study considers the output lost due to the loss of what the individual would have produced during the year of his or her death.60 They calculate this 57Intentional violence cases accounted for 39.4 percent of cases at emergency rooms at the University Hospital of the West Indies in 1996 and 43 percent of emergency room visits at public hospitals in 2001. (McDonald et al., 1999) and Francis et al. (2003). 58An earlier study, Mansigh and Ramphal (1993) examined the costs of treating interpersonal violence in Kingston Public Hospital and estimated them at a value equivalent to $709 in 2001 U.S. dollars. 59The study values the time of injury victims using the average weekly wage estimated in the study for murder victims of J$8423. While it is not clear in the study, presumably this average is among those who are employed. Because unemployment is high in Jamaica, this is likely to be an over-estimate of the value of victims' time in terms of wages. If not incapacitated by injury, some victims would be unemployed or out of the labor force rather than working. 60Note that it would be preferable to calculate the net present value of a discounted lifetime stream of wages, rather than take the wages in one year. The use of just one year's wages gives an underestimate of the future value of the individual's production. Calculations like this of the value of a person's life are controversial because they essentially assume that the life of a person with low-earning power is worth less than that of someone who earns more. 51 by examining the occupational distribution of a random sample of murder cases in the previous three years and constructing a corresponding average wage. Multiplying this wage (on an annual basis) by the number of murder victims, they find a total lost output of J$194 million. 4.34 Security costs dominate the total costs of crime as calculated by the accounting method. In the year examined in the study, the Government's budget for security services, including defense, justice, correctional services, and the police, totaled J$10.5 billion or about 3.1 percent of GDP. Government expenditures on the justice system and the police are oriented chiefly towards criminal cases; of all cases filed with the Resident Magistrate courts in 2001, just 10 percent were civil cases. For the purposes of the costs exercise the full value of public security forces is counted as a cost of crime. Table 4.2: The Costs of Crime in Jamaica: an Accounting Exercise 1) Health Costs 1.3 bn (0.4% of GDP) Public Health System 995.7 Private Citizens 254.5 2) Lost Production 0.5 bn (0.2% of GDP) Mortality 194.1 Injury Due to Crime 337.2 3) Public Expenditure on Security 10.5 bn (3.1%of GDP) Total (1) + (2) + (3) 12.4 bn (3.7% of GDP) Source: Francis et al. (2003). 4.35 Table 4.2 above summarizes the costs of crime as calculated by Francis et al. (2003). They find that the total costs of crime in 2001 came to J$12.4 billion, which was 3.7 percent of GDP. 4.36 It is important to note what costs are not included in the 3.7 percent of GDP figure. First, this does not include the private security costs to firms shown in Figure 4.7, which average 2.0 percent of firm revenue across all firms.61 The Francis et al. study makes no attempt to quantify non-monetary costs like the pain and suffering of victims and their families and the psychological effects of living in fear of being victimized. The estimate of crime's effect on GDP also does not take into account the long-run economic or social effects of violence, such as the impact of violence on capital accumulation and future growth rates. In particular, there is a risk of a vicious circle, where violence- plagued neighborhoods receive little productive investment and hence offer few productive employment opportunities. The lack of opportunities, in turn, could lead youth to engage in violent and criminal activities. Nor does the cost estimate capture the fact that children who grow up in violent environments are themselves more likely to be violent as adults. This last effect--sometimes called "the inter-generational transmission of violence"--means that violence has significant inertia. 61Because the firm survey is not representative of the population of Jamaican firms and because security costs as a percentage of revenue vary greatly with firm size, it is not possible to reliably convert the 2.0 percent of revenue figure into a percentage of GDP. 52 DISABILITY-ADJUSTED LIFE YEARS LOST TO VIOLENCE 4.37 Another way to summarize the costs of violence is by examining the Disability- Adjusted Life Years (DALYs) lost to violence. DALYs are the standard international health measure of the burden of disease. DALYs express years of healthy life lost, summarizing both the years of life lost to mortality and the suffering or incapacity associated with morbidity (Mathers et al., 2003). An important strength of analysis based on DALYs is that it provides a clear basis to compare the impact of violence on health with impacts from other threats to health. 4.38 The World Health Organization (2002) publishes estimates of cause-specific deaths and Disability-Adjusted Life Years (DALYs) lost, and violence is among the listed causes. 62 One weakness of the WHO figures is that they measure only the direct effects of violence on those who are killed or injured and do not capture indirect effects on other individuals (such as increased stress levels or other reductions in the quality of life due to the fear of violence) or on society more broadly (such as lowered saving, investment and growth). 4.39 The DALYs lost to violence are calculated based on two estimated components: age-specific rates of death (mortality) due to violence, and age-specific rates of suffering and incapacity (morbidity) due to violence. The WHO publishes the overall deaths due to each cause, including violence. 4.40 The quality of the WHO data varies enormously by country. The reliability of the WHO data can be considered by examining the WHO figures on deaths due to violence with official homicide rates for each country. Figure 4.8 shows a scatter plot of the WHO data for 2002 vs. data from official sources for the closest available comparison period. The official source figures for Jamaica, the Dominican Republic, and Guyana are the respective governments' published figures for 2002. The Haiti figure is calculated from responses to a 2001 household survey. Figures for other countries are from government reports for various periods during the 1990s. 4.41 The figure shows that for particular countries the WHO data are substantially in error. In the case of Jamaica, in particular, the WHO data appear to be significantly flawed: the WHO data for 2002 show a murder rate of 0.5 murders per 100,000 Jamaicans, compared to the rate of 40 per 100,000 reported by the police for that year. Likewise, the WHO's figure (10.2) for Haiti is less than a third of the estimate based on 2001 household survey data (33.9).63 For Guyana and the Dominican Republic--the two other large countries in the region for which both official and WHO data are available-- the WHO figures are also substantially below official homicide rates. Nonetheless, excluding the two cases where the data are clearly deeply flawed--Jamaica and Haiti-- the WHO figures do not appear to show any systematic bias with respect to the official figures. For the remaining eight countries, the WHO estimate is above the official figure 62Jamaica homicide figures from multiple sources are examined in detail in the World Bank's forthcoming Jamaica Poverty Assessment (2006). 63Note that the household survey-based estimate for Haiti has a wide confidence interval and is not as reliable as the estimates from official reports. 53 in half the cases and below it the other half. These can be seen in the figure as the points that fall above and below the 45 degree line. This suggests that although there are problems with the data, for countries other than Jamaica and Haiti, the DALY figures are not necessarily over- or under-stated and may be cautiously interpreted as a measure of violence's cost in terms of health. Figure 4.8: How Reliable is WHO Homicide Information for the Caribbean? Homicide Rates According to WHO vs. Official Sources Homicide Deaths per 100,000 Population 40 Jamaica se 35 Haiti Sourclaci 30 Offi to gnidro 25 Acc 20 2002, Guyana ac 15 Dominican Republic 45 Cir degrees St. Kitts & Nevis Rate, 10 DominicaBarbados Trinidad & Tobago deci Grenada 5 Antigua & Barbuda Homi 0 0 5 10 15 20 25 30 35 40 Homicide Rate 2002, According to WHO Global Burden of Disease Database Source: WHO and various national sources. 4.42 Figure 4.9 shows the DALYs lost to violence by country in the Caribbean, according to WHO figures, both in raw totals and per 100,000 residents. The largest numbers of DALYs lost to violence are in the region's largest countries: Cuba, the Dominican Republic, Haiti, and Guyana. In terms of population-standardized rates of DALYs lost, Guyana ranks at the top in the WHO figures, followed by St. Vincent and the Grenadines and Saint Kitts and Nevis. It should be noted, however, that because the populations of the OECS countries are small, the population-standardized rates are sensitive to small errors in the DALY figures. 54 4.43 The rate of DALYs lost to violence in almost all Caribbean countries is above the worldwide average. The exceptions are Dominica, and the two countries for which the data is evidently in error: Haiti and Jamaica. Figure 4.9: Disability-Adjusted Life Years Lost to Violence in the Caribbean, 2002 Total DALYs (in 1000s) DALYs per 100,000 population Cuba 44.3 Guyana 716 Dominican Republic 40.2 Saint Vincent and the Grenadines 624 Haiti 25.1 Saint Kitts and Nevis 534 Guyana 5.5 Dominican Republic 467 Bahamas 3.0 Saint Lucia 458 Barbados 1.1 Antigua and Barbuda 431 Jamaica 1.0 Barbados 416 Saint Vincent and the Grenadines 0.7 Cuba 393 Saint Lucia 0.7 World Average 343 Antigua and Barbuda 0.3 Haiti 306 Saint Kitts and Nevis 0.2 Dominica 238 Dominica 0.2 Jamaica 37 0 10 20 30 40 50 0 100 200 300 400 500 600 700 800 Source: WHO (2002). Note: The world average DALYs per 100,000 population is calculated across all countries, weighted by population. Figure 4.10: Disability-Adjusted Life Years Lost to Violence vs. Other Causes Caribbean, 2002 Carib. Average (Weighted) Carib. Average (Weighted) Carib. Average (Unw eighted) Violence Carib. Average (Unw eighted) Antigua and Barbuda Antigua and Barbuda Bahamas Bahamas Violence Barbados Communicable, maternal, Barbados Cuba perinatal and nutritional Cuba HIV/AIDS Dominica conditions Dominica Dominican Republic Dominican Republic Guyana Guyana Haiti Haiti Jamaica Jamaica St. Kitts and Nevis St. Kitts and Nevis St. Lucia St. Lucia St. Vincent and the Grenadines St. Vincent and the Grenadines 0 5000 10000 15000 20000 25000 30000 35000 0 2000 4000 6000 8000 10000 Carib. Average (Weighted) Carib. Average (Weighted) Carib. Average (Unw eighted) Carib. Average (Unw eighted) Antigua and Barbuda Antigua and Barbuda Bahamas Bahamas Barbados Barbados Violence Cuba Cuba Dominica Dominica Unintentional Dominican Republic Violence Dominican Republic injuries Guyana Guyana Haiti Noncommunicable Haiti Jamaica diseases Jamaica St. Kitts and Nevis St. Kitts and Nevis St. Lucia St. Lucia St. Vincent and the Grenadines St. Vincent and the Grenadines 0 2000 4000 6000 8000 10000 12000 14000 16000 0 500 1000 1500 2000 2500 3000 Source: WHO (2002). 4.44 How big is the direct impact of violence on DALYs relative to other causes? Overall, relative to all other causes, the impact seems to be relatively small, although this conclusion must be tentative because of the data problems affecting several countries in the region. In the WHO data, violence account for just 0.37 percent of all DALYs lost 55 and 0.90 percent of all deaths in 2002 in the Caribbean. Figure 4.10 shows graphically DALYs lost to some of the other major categories. Other major health conditions, including HIV/AIDS, dwarf violence in terms of their impact on DALYs. For the Caribbean overall, DALYs lost to violence are equivalent to 24 percent of those lost to unintentional injury.64 As a whole, the DALY figures suggest that the direct health impact of violence--at least as measured in the WHO figures--may not be the most important part of the effect of crime and violence at a societal level. THE IMPACT OF VICTIMIZATION ON SELF-REPORTED LIFE SATISFACTION 4.45 Another cost of crime and violence is the effect being a victim has on one's quality of life. It is well known that exposure to crime can have long-lasting psychological impacts on the victims and those close to them. However, very little empirical research has been conducted on the relationship between crime and well- being.65 We analyzed the effect of being a victim or having a family member be a victim on subjective well-being, using Latinobarómetro surveys from 18 countries in Latin America.66 The only Caribbean country included in the survey is the Dominican Republic. 4.46 In the Dominican Republic, controlling for a wide variety of personal and socioeconomic characteristics, those who have been victimized (or had family members victimized) in the previous 12 months are 8 percent less likely to say that they are "satisfied" or "rather satisfied" with life. The effect in the Dominican Republic is higher than the average estimated effect of 4.4 percent for the 14 countries surveyed. Full results from the analysis are shown in Annex 4.1. 4.47 It is possible that this reflects in part not a causal effect of victimization but rather the fact that victim and non-victim families differ along other lines relevant to self- satisfaction. For example, it may be that victimized families tend to live in particular neighborhoods with characteristics that make people both less likely to be satisfied with life and more likely to be victims of crime. The long list of covariates used in the analysis controls for such differences as best as possible using the available data. Overall, the analysis presented here provides reasonably strong evidence that crime has a substantial impact on the quality of life of victimized families and that this effect may be particularly high in the Dominican Republic. 64 The WHO divides injuries into two categories: intentional and unintentional. Subcategories for intentional injuries are self-inflicted injuries, those due to war, and those due to violence. The subcategories for unintentional injuries are as follows: road traffic accidents, poisonings, falls, fires, drownings, and other. 65One exception is Powdthavee (2005), which examines crime and subjective welfare data from a 1997 survey in South Africa. 66The relevant questions in the survey are "In general terms, how satisfied would you say you are with life? Very satisfied, rather satisfied, not very satisfied, or not satisfied at all?" and "Have you or someone in your family been assaulted, attacked, or the victim of a crime in the last 12 months?" 56 THE IMPACT OF VIOLENT CRIME ON ECONOMIC GROWTH 4.48 Another way to assess the costs of crime is to estimate the impact of crime on overall economic growth using cross-country panel data.67 The advantage of this approach is that it summarizes the overall economic cost of crime. Because the estimates are based on data from across countries, they reflect an average relationship between crime and growth worldwide. This section reviews estimates produced using this method as part of a study of crime and violence in Brazil (World Bank, 2006b) and considers their implications for Jamaica. 4.49 In this analysis, violent crime rates are measured using national homicide rates. Homicide rates are typically used for cross-country crime studies because they are thought to be least subject to variation in definition and reporting to authorities. The analysis follows the literature on the determinants of growth: GDP per capita is regressed on homicide rates, controlling for a country's level of income inequality, the cost of investment, and average male and female education. 68 4.50 The measure used for this analysis is "completed homicides", referring to actual homicides.69 The analysis employs the Arellano and Bond (1991) GMM estimator, which provides consistent estimates of the effect of the explanatory variables on the outcome under the identifying assumption that the explanatory variables are not correlated to the time-variant components of the error terms. In all cases, the crime data is averages over five-year periods between 1975 and 2000.70 4.51 Estimates are presented in Table 4.3. The dependent variable is the log of future GDP per capita, while "Income" as shown in the table is the log of current GDP per capita. The coefficient estimate on the homicide rate is significant and negative. This indicates that a country's homicide rate has a negative impact on subsequent economic growth71 . 67 Obviously, causality runs in both directions: violence affects growth, but growth--by influencing opportunities for gainful employment--also affects violence. Here, we focus on the causal arrow running from violence to growth. See Chapter 3 for a more complete discussion of these causality issues. 68 Data on homicides for 1975-2000 is drawn from the United Nations Survey of Crime Trends and Operations of Criminal Justice Systems. Data on schooling comes from Barro and Lee (2000). Income inequality data comes from Deininger and Squire (1996). 69This definition may seem tautological, but international data distinguish between completed and total homicides. Completed homicides are actually consummated homicides, while total homicides include both consummated and attempted homicides. 70Dependent variable is average annual per capita growth. Robust standard errors are in parentheses. A constant was included in the model, as were dummies for time period. Observations are for five-year time periods between 1975 and 2000. Note that in this analysis violent crime rates are measured using national homicide rates. Homicide rates are typically used for cross-country crime studies because they are thought to be least subject to variation in definition and reporting to authorities. The analysis follows the literature on the determinants of growth: GDP per capita is regressed on homicide rates, controlling for a country's level of income inequality, the cost of investment, and average male and female education. 71The regression estimates imply that violent crime substantially reduces economic growth. A decline of 10 in a country's completed-homicide rate per 100,000 persons is associated with a 2.1 percent increase in average annual growth over the next five years. 57 Table 4.3: Cross-Country Regression Estimate of the Effect of Violent Crime on Economic Growth Variable Estimates Income -0.1362 (.017) Inequality 0.0013 (.0005) Male education 0.0120 (.0168) Female education -0.0084 (.0179) Price level of investment 0.0000 (.0001) Homicide rate -0.0021 (.0006) Period Dummies Yes Countries 28 Observations 32 Source: World Bank (2006b). Note: Standard errors are shown in parentheses. 4.52 We can consider what these estimates imply for potential gains to economic growth from crime reduction in the Caribbean. For 1996-2000, Costa Rica--one of the least violent countries near the Caribbean--had a homicide rate of 8.1 per 100,000 population according to data collected by the UN Office of Drugs and Crime. In contrast, the homicide rates in Jamaica, Haiti, Guyana and the Dominican Republic for the nearest available comparable period were 33.8, 33.9, 16.1, and 16.5, respectively.72 What would be the gain in economic growth for these Caribbean countries if they were to bring down their homicide rates to that of Costa Rica? Estimates from this "simulation" are shown in Figure 4.11. 72The Jamaica homicide rate figure is the average over 1998-2000 from the UN database. Because recent UN data is not available from the other three countries, homicide rates for those countries are drawn from alternative sources and for varying dates: 2001 household survey (Haiti), and the averages of 1999-2004 police reports (both Guyana and the Dominican Republic). 58 Figure 4.11 Potential Boost to Annual Economic Growth Rate from Reducing Homicide Rate to Costa Rica Level thworG 6% 5.4% 5.4% e 5% moc In 4% itapaCrePlaunnA 3% 2% 1.8% 1.7% in esaerc 1% In0% Dominican Guyana Haiti Jamaica Republic Source: Own calculations. 4.53 The regression results suggest very large potential gains from reduction in violence for Haiti and Jamaica, the two countries in the region with the highest murder rates. The coefficient estimate implies that both countries could boost economic growth per capita by 5.4 percent per year if they were to bring their homicide rates down to the levels of Costa Rica. Guyana and the Dominican Republic would also benefit substantially, with potential growth rate increases of 1.7 percent and 1.8 percent, respectively. 4.54 Because the estimated effects are on annual growth, they are cumulative over time, suggesting that over the long term the impact of crime reduction on welfare would be very high. For example, an increase of per capita growth of 1.8 percent--the figure implied by the simulation for the Dominican Republic--would result over 20 years in a cumulative increase of income per person of 43 percent. POLICY IMPLICATIONS 4.55 There are a number of ways to consider the costs of crime. Each approach illuminates a different aspect of the impact of crime, and no single approach is comprehensive. Even the accounting approach, which attempts to pull together a large variety of costs in a single measure, is limited to the extent that some costs of crime cannot be quantified; it is also generally limited to the current costs which can be more easily identified, rather than longer-term impacts. 4.56 The work reviewed and presented in this chapter shows that some of the most important costs such as impact on the investment climate and economic growth are those that are usually neglected in many analyses and in public discussions. Most importantly, 59 the economic growth effects implied by the dynamic panel estimates are very large. They suggest that the countries of the Caribbean, which suffer from the most severe crime and violence problems, could realize a very large dividend in increased growth from reducing them. 60 5. YOUTH VIOLENCE IN THE CARIBBEAN: A CASE STUDY OF THE DOMINICAN REPUBLIC Deaths and injuries from youth violence constitute a major public health, social and economic problem across the Caribbean, including in the Dominican Republic. Youth are overrepresented in the ranks of both victims and perpetrators in the Dominican Republic, and this pattern has become more stark over time, as rates of crime and violence overall have increased. A wide variety of risk factors contribute to the prevalence of youth violence, including poverty, youth unemployment, urban migration, drug trafficking, a weak education system, ineffective policing, the widespread availability of weapons, drug and alcohol use, and the presence of organized gangs. Nonetheless, youth violence is preventable; a broad range of viable strategies for preventing and reducing youth violence exist. Most highly effective programs combine components that address both individual risks and environmental conditions, building individual skills and competencies, training parents for greater effectiveness, improving chances for poor youth to access and complete their secondary education, improving the social climate and safety of school, and providing "second chances" to those who have dropped out of the formal schooling system, including school equivalency programs, job and life skills training, and apprenticeships. 5.1 Youth violence is a high-priority, high-visibility concern across the Caribbean. Not only has violence grown in most of the region in recent decades, but youth are also disproportionately represented in the incidence and severity of this trend, both as victims and as perpetrators. Moreover, violent crimes are being committed at younger ages in many countries. Yet there is growing evidence that youth violence can be prevented and offenders can be rehabilitated when appropriate policies and interventions are adopted (WHO, 2002; US Department of Health and Human Services, 2001; World Bank, 2005c; IADB, 2002; Tolan and Guerra, 1994). This chapter addresses the pressing issue of youth violence and its implications for the Caribbean, with a particular focus on the Dominican Republic. The chapter also provides a set of policy and program recommendations which could prove useful to the region as a whole. 5.2 The overarching conclusion is that although youth violence is a legitimate concern for the region, it is neither intractable nor are youth "the problem." Rather, as detailed by the World Bank's Caribbean Youth Development Report (World Bank, 2003a), youth are a product of a complex set of factors in their environments. They also represent a unique window of opportunity to both prevent and reduce crime and violence in society at large. Evidence from evaluated youth violence prevention programs outside the Caribbean indicates that the earlier the investment in an individual, the greater the chance that violent behaviors can be prevented through adulthood, and the more cost- effective the investment (WHO, 2003; World Bank, 2005c; Schweinhart, 2005 and Levitt, 1998). There are a multitude of policies and programs currently underway across the Caribbean to address youth violence. Unfortunately, very few have been subjected to rigorous impact evaluations and, consequently, there is little region-specific knowledge about what works and what does not. However, there is increasing evidence globally on 61 the types of youth violence prevention efforts that work, and many of these programs are already being implemented in the sub-region. SCOPE OF YOUTH CRIME AND VIOLENCE 5.3 This section provides a brief overview of youth crime and violence occurring in the Caribbean, with a special focus on the Dominican Republic. For the purposes of this report, youth violence is defined as homicide and non-fatal attacks perpetrated by or against a person aged 10-29 years of age.73 This age range was selected for three reasons. First, the late adolescent/early adulthood period of 15-29 is generally considered to have the highest risk for all forms of violence globally, and is particularly so in the Caribbean. Second, the early adolescent period (10-14 years) is a growing concern in the region because both quantitative and qualitative evidence points to violent crimes being committed at younger and younger ages. While still considered to be one of the lowest risk groups across the entire lifespan for being a victim of homicide, young adolescents are observed to be increasingly involved in both homicide and other forms of crime and violence. More importantly, they are the age at which many prevention policies may have a greater chance of success. Finally, the age range 10-29 coincides with that used in the only globally comparable data available on youth violence: World Health Organization statistics.74 5.4 Data Limitations. As noted elsewhere in this report, general crime and violence data is often difficult to obtain and are often incomplete (see Chapter 1). This challenge becomes even greater when attempting to obtain information on youth in particular. The problem stems from a lack of common and comparable definitions of youth when it comes to crime and violence, as well as weak systems for surveillance and monitoring across what are considered minor (under 18) and adult (18 and older) age ranges. In particular, when attempting to compare data across countries, there are very few indicators beyond homicides that are disaggregated by comparable ages, and the Caribbean has particularly weak youth data compared with much of the rest of Latin America. This report has drawn from both available international quantitative sources (WHO, 2002/2003; PAHO; UNODC) as well as from primary quantitative and qualitative sources in the Dominican Republic (ALEPH, 2006). 5.5 Youth are disproportionately represented in Caribbean crime and violence, both as victims and as perpetrators, and in many of these countries violent crime is being committed at younger ages than in the past. Moreover, there is a growing concern that in a number of countries, the proportion of violent crimes committed by youth, and especially juveniles, has been increasing (UNODC, 2002). In 2005, young Dominicans aged 11-30 accounted for 46 percent of homicide victims, yet only represented 38 percent of the general population. In Jamaica, youth under the age of 25 were responsible for 51 percent of all murders and 56 percent of all major crimes in 2000 (World Bank, 2004). In the Dominican Republic, arrests for homicides by minors under the age of 18 rose over 73The 10-29 age range thus includes youth and young adults. 74Note that the age range will depend largely on available data. Much of the standardized WHO data on youth violence use the 10-29 age range, whereas the official UN MDG definition of youth is 15-24 years old. 62 the period 1995-2004 from 2 to 113, with over 95 percent male (UNODC, 2002). Similar trends occurred in St. Kitts and Nevis, where in 1990 only 1.2 percent of all crimes were committed by juveniles, yet by 1998, this had increased to 17 percent (UNODC, 2002). A worrisome consequence of these trends is that evidence points to the fact that violent behavior in youth has a strong tendency to continue into adulthood (WHO, 2003; Levitt, 1998). 5.6 Youth as victims. How do youth violence statistics in the Caribbean compare with other countries? Youth homicide rates globally have large variations: for males, the numbers range from 2.5 per 100,000 in Canada and 5.2 in Chile, to 94.8 in El Salvador and 156.3 in Colombia (WHO, 2002). The Latin America and the Caribbean region boasts the highest homicide rate of men between the ages of 15-29 (68.6 per 100,000) in the world, more than three times greater than the global average of 19.4 (Table 5.1). 5.7 In the Dominican Republic in 2002, the adjusted homicide rate was 19.7 (per 100,000) for young Dominicans aged 15-29, compared with 10.2 for all Dominicans (Aleph, 2006).75 As Figure 5.1 illustrates, homicide deaths for youth have been growing consistently since 2000, particularly in the 18-29 age range. In 2005, homicides of those ages 11-30 accounted for approximately 46 percent of total homicide deaths.76 Young males are particularly affected, with a death rate of 35.3 (per 100,000 inhabitants) in 2002, which is over 80 percent higher than the global average of 19.4 for young men (WHO, 2002). They are also considerably more likely to be victims of homicide than the rest of the Dominican male population (18.3); moreover, they are fifteen times more likely to die from homicide than women as a whole (2.1 per 100,000 inhabitants) and nearly 8 times more than women of the same age group (3.2) (ALEPH, 2006; PAHO, 2006). 5.8 Violence against young women. Though often different in nature,77 violence against women is also disproportionately borne by youth in the Caribbean. The most widespread type of violence in the Caribbean is that which occurs within families and intimate relationships, where girls and young women are disproportionately affected, especially if the violence involves sex (PAHO, 2000). See chapter 1 for more on violence against young women. 75 Data obtained from the Procuraduría General Statistics Department. 76 Ibid. Results of self-report surveys around the world indicate that an overwhelming majority of those who participate in violence against young people are about the same age and gender as their victims; in most cases the offenders are males acting in groups (United Nations, 2003). 77 Violence against women was defined by a declaration of the General Assembly of the United Nations (Declaration on the Elimination of Violence against Women, 1993) as "any act of gender-based violence that results in, or is likely to result in, physical, sexual or mental harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life." 63 Table 5.1 Youth as Victims: Homicide Rates in Select Countries (per 100,000 inhabitants) All Country Year Total Ages All Ages Male Female Males Female Age15-29 Age15-29 Bahamas 1995-1997 14.9 26.1 -- 48.4 -- Brazil 1995a 16.7 23.0 4.1 81.2 6.5 Chile 1994 3.0 5.4 0.8 6.7 -- Cuba 1997 6.2 9.6 2.7 18.4 5.7 Colombia 1995 61.6 116.8 9.0 212.5 15.0 Dominican 2002a 10.2 19.7 2.12 35.3 3.22 Republic 2005b 26.41 49.9 4.4 -- -- El Salvador 1993 55.6 108.4 8.4 133.1 8.8 Guyana 1994-1996 6.6 11.8 -- -- -- Jamaica 2004a 55.7 102.1 10.5 188.0 14.8 Trinidad and Tobago 1994 12.1 17.1 6.6 21.6 -- LAC average 19.3 34.7 4.0 68.6 6.4 United States 1998 6.9 10.7 3.1 23.6 4.6 Canada 1997 1.4 1.9 1.0 3.2 1.1 World 8.8 13.6 4.0 19.4 4.4 Source: World Health Organization. 2003. World Report on Health and Violence. Geneva. a: Based on data obtained from World Health Organization (August 2006). b: Estimates from National Police Statistics. 5.9 Youth as perpetrators. Just as they account for a disproportionate share of the victims of violence, young people are also disproportionately its perpetrators, especially young men. In most countries, this is a growing trend. Indeed, statistical data indicate that in virtually all parts of the world, with the exception of the United States, rates of youth crime rose in the 1990s, with many of the criminal offenses related to drug abuse and excessive alcohol use (UNDP, 2003). What little data exist indicate that this phenomenon may be particularly worrisome in the Caribbean. 64 Figure 5.1: Homicide Deaths by Age (10-29) in the Dominican Republic (2000-2005) 100 90 80 2005 shat 70 2002 defo 60 erb 50 2000 40 Num 30 20 10 0 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Age Source: Dominican Republic National Police, unpublished data. 5.10 In Jamaica, for example, the country's homicide rate for 2005, considered the highest in the world, was at an all-time high and over 70 percent of homicides were committed by young men ages 16-30. Shares were similar for shootings (78 percent), robbery (74 percent), and rape (68 percent). In the Dominican Republic, of the current prison population, 62 percent of those arrested for homicide were 16-29 at the time of arrest compared to 71 percent of those arrested for robbery. In the first six months of 2006, 61 percent of new prisoners arrested for homicide were 16-29 and 70 percent for robbery (Procuraduría General de la República Dominicana, 2006). 5.11 Crimes perpetrated by minors (under age 18) in the Dominican Republic have also been on the rise over the past decade. Arrests of minors tripled between 2000 and 2003 and started to decline in 2004 and 2005 (Figure 5.2). However, those arrests related to illegal arms and drugs have shown no sign of decreasing (Dominican Republic National Police, 2006) In fact, whereas over the past decade the most common crimes committed by minors were petty theft (48 percent), drugs (11 percent) and assault (10 percent), those that experienced the most consistent growth were arrests for homicide and illegal arms, with average annual increases over the past decade of 195 percent and 107 percent, respectively. These were followed by average annual increases in aggressive assault (28 percent) and drug arrests (22 percent).78 As Box 5.1 illustrates, qualitative data supports the quantitative evidence that younger groups of youth are increasingly involved in violent crime. 78Author's calculations based on Dominican Nacional Police statistics (1995-2005). 65 Figure 5.2: Trends of Minor Arrest over Past Decade in the Dominican Republic 3500 Illegal Arms Rape-Sexual Violation Aggression 3000 Homicide Fighting Drugs rso 2500 Petty Theft Minfo s st 2000 re Arfo er 1500 mb Nu 1000 500 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year Source: Dominican Republic National Police Crime and Victim Data, 2006. 5.12 Youth focus groups reveal some common characteristics of youth perpetrators of crime. The overwhelming majority of youth currently residing in the Dirección Nacional de Centros de Atención Integral (juvenile corrections center) are young men (136, compared with 3 women). The primary motivation identified for committing a crime was a desire to escape poverty (ALEPH, 2006). Other common characteristics were: growing up in dysfunctional households, having been abused and mistreated, getting an early start in a life of crime (with a reported average age of first crime as 13), consuming illegal drugs, and having dropped out of school. 5.13 Gang and drug related violence has also increased in recent years, with youth as the most visible culprits of this type of crime and violence. Because the criminal justice system in the Dominican Republic does not allow arrest or incarceration of youth below the age of 13,79 gangs and drug lords are increasingly using younger members to carry out both petty and hard crimes (Box 5.1). This appears to be a trend throughout the sub- region (see discussion of gangsterism in next section). As reported in the World Development Report (WDR) 2007, younger gang members are responsible for a disproportionately large share of offenses. The WDR finds that the formal association with a gang is powerful: gang members wield more influence on the violent behavior of their peers than violent non-gang members. What is more, youth gang members tend to commit crimes that are more serious and violent while they are gang members than after they leave the gang (World Bank, 2006c). 79The "Ley 136-03, Código para el Sistema de Protección y los Derechos de Niños, Niñas, y Adolescentes," article 223 states that children under the age of 13 are in no case to be held criminally responsible, and as such, cannot be held for arrest, nor be given any sanction by official authorities. See Guerrero (2005). 66 Box 5.1: The Sophisticated Extortion of Santo Domingo Gangs and Drug Lords There is emerging quantitative and qualitative evidence that younger adolescents are increasingly involved in violent crimes, particularly as they relate to the drug trade: "Never in the past have we witnessed such sophisticated tactics used by drug lords and young gang members to attract children to carry out their crimes. Over the past three years, we have seen one particularly cruel method that has taken on a frightening momentum, and its impact on our society's children is devastating." "Young men who we know to be involved in the drug trade (because we have lived in their neighborhood for years) invite young boys--usually 8-12 years old from broken families where the mother has to work--to join in sports activities and games in the neighborhood. These young men have nice clothes, shoes, cell phones, and they become buddies with the children." "Once they gain their trust, they ask them who they love most in the world. The children respond, and are then told, if you do not steal 3 cell phones a day for us, then we will kill your mother' (or whomever else they have confided is their most beloved person). A 10 year-old child from the barrio has nothing else to hold on to. We have witnessed in the past 6 months hundreds of crimes committed this way ranging from petty theft to murder. These children never see a way out...." - Interview with a group of Catholic nuns who have served in several of Santo Domingo's most violent neighborhoods for over 15 years. May 20, 2006 FACTORS CONTRIBUTING TO YOUTH VIOLENCE 5.14 Identification of the key risk factors that contribute to youth violence makes clear that the policy and programmatic responses need to extend across a wide range of professional disciplines, including those of public health, education, and skills training. This section summarizes the current literature discussing these factors with a view to informing policymakers on where to most effectively target interventions. As outlined in Chapter 3, risk factors can be identified at four interrelated levels: i) societal; ii) community; iii) relationships; and iv) individual. In addition to "risk" factors, protective factors are also present at each of the four levels, decreasing the likelihood that youth engage in crime or violence. It is worth emphasizing that the presence of a risk (or protective factor) does not ensure that criminal or violent behavior will (not) occur--it merely increases the likelihood that it will (not). While the analysis here focuses on factors for youth violence in the Dominican Republic, where possible the findings will be placed in a wider Caribbean context. 5.15 Recent empirical research conducted on risk and protective factors for youth in the Caribbean concluded that school attendance/connectedness are the most important factors in reducing violent youth behavior. In particular, the study found that boys (girls) who feel connected to school were 60 percent (55 percent) less likely to engage in violent activity.80 In addition, the study showed the significant effects that schools have in reducing drug use, smoking, and alcohol consumption. The study also found that family 80Analyses were done on the results of a 1997-98 survey of over 15,500 young people 10-18 years of age in 9 Caribbean countries to identify risk and protective factors associated with health compromising behaviors, including violence. See Blum, R.W. et al. (2003). 67 connectedness, or the presence of a caring adult, served as the second most important protective factor. The analysis concluded that both risk and protective factors are cumulative; if protective influences are held constant and predominant risks are added one at a time, risk behavior rises significantly. Conversely, and perhaps more importantly, when risk factors are held constant and protective factors are added, there is an even greater reduction in reported involvement with violence. 81 5.16 Societal Level: Some examples of risk factors at this level are poverty, inequality, youth unemployment, an ineffective criminal justice system, and drug trafficking. At the societal level in general, the Caribbean has had a culture of aggression rooted in slavery since colonial times. Today, messages to youth from both regional and international (U.S.) media and music in particular, are often expressions of rage or alienation--anti-women, pro drugs, pro-violence, and materialistic--that influence the decisions that youth make (World Bank, 2003a). For the Dominican Republic, some of the identified trends at this level are as follows: 5.17 Poverty and Inequality. Being raised in poverty has been found to contribute to a greater likelihood of involvement in crime and violence. It is also often related to youth aggression because of increased stress and feelings of hopelessness that may arise from chronic unemployment and other associated factors (IADB, 2002; Weaver and Maddeleno, 1999). The inverse relationship between family income and juvenile crime is well documented globally (World Bank, 2006d). During the deep economic and financial crisis of 2003-2004, extreme poverty doubled from 7 percent to 14 percent (World Bank and IDB, 2006). Violent crimes also rose dramatically; from 2002 to 2005, the rate of violent death nearly doubled from 14.5 to 26.4 per 100,000 residents (Aleph, 2006).82 In addition to poverty, the income inequality demonstrated by drug dons, foreign tourists, and the media encourages engagement in easy money activities, including drugs and prostitution (World Bank, 2003a). 5.18 Youth Unemployment. Between 2002 and 2004 youth unemployment jumped from 23 to 31 percent, or more than twice the LAC average (15.2 percent) and the global average (14.4 percent) (World Bank and IDB, 2006). Although data is not available, it is generally agreed that these figures for urban poor youth are significantly higher. The fact that young Dominicans in (or entering) the labor market are at a disadvantage is consistent with findings from the international literature on the effects of demand shocks (Jimeno, J. and Rodriguez-Palenzuela, D, 2002), as well as those observed throughout the Caribbean, where many countries have especially elevated youth unemployment. International comparisons indicate that Barbados, the Dominican Republic, Jamaica, St. Lucia and Trinidad and Tobago have high rates compared to the rest of Latin America (Table 5.2). Evidence from the United States has shown that falling wages and local youth unemployment were partly responsible for the rise in youth crime in the 1970s and 1980s (World Bank, 2006d). 81Ibid. 82Note that violent deaths in the Dominican Republic include homicides and deaths resulting from shootings between police and civilians. In 2005, 18 percent of violent deaths were a result of shootings between police and civilians; the remainder were homicides. 68 Table 5.2: Highlights of Youth in Numbers in Latin America and the Caribbean Youth Percentage of Net Fertility Country Youth Secondary Unemployment youth population (15-24) School Enrolment (15-19) Bahamas 15.8 18.0 79 49 Barbados 21.8 15.1 87 46 Dom Rep (2002) 23.1 19.1 41 94 Dom Rep (2004)a 31.0 19.1 38 N/A Grenada 27.0 24.5 -- 69 Jamaica 34.0 18.7 75 32 St. Lucia 44.0 20.3 70 46 St. Vincent and the Grenadines 36.0 20.8 -- 22 Trinidad and Tobago 25.4 20.7 72 13 LAC Average 15.2 -- -- 52 United States 10.6 14.2 85 57 Source: World Bank 2004 , World Bank 2005d, and Secretaría de Educación. 5.19 Urban migration is on the rise, with a high share of young migrants who are "unattached" to families, schools or employment. Although large youth cohorts in and of themselves are not always associated with increases in crime and violence, the correlation is stronger when taken in the context of increasing poverty and rapidly growing cities, such as in the urban areas of the Dominican Republic (World Bank and IDB, 2006). Migration flows in the Dominican Republic are very high, with more than 2 million people living in a region other than that of their birth in 2002 (World Bank and IDB, 2006). Whereas 56 percent of Dominicans were living in cities in 1997, that number had reached 64 percent by 2004 (Fares et al., 2006). Nearly 40 percent of Dominicans choosing to resettle are between the ages of 20 and 39 (Fares, et al., 2006). 5.20 Drug trafficking. The Dominican Republic's role as a transit country for the drug trade puts many young Dominicans at risk. This macro level trend disproportionately affects youth in two ways. First, given the immunity of minors from prosecution in adult courts, the relative cost of their involvement in the drug trade is lower than that of their adult counterparts, giving them a perverse comparative advantage. Second, remuneration in the narcotics business tends to be in-kind, and this trend is 69 growing in the Dominican Republic, particularly with younger members. 83 This creates a long-term involvement with the industry of illicit drugs, expanding the client base for the drugs and creating a dependency/addiction among participating youth. 5.21 Community level: Some common risk factors at this level relate to schools, neighborhoods, and police. Examples are lack of school access, school policies, availability of guns, prevalence of drugs, unsafe neighborhoods, police abuse and criminal justice responses at the local level. The presence of these factors in the lives of young Dominicans is as follows: 5.22 Schools. The education sector in the Dominican Republic is characterized by low enrollment rates, high dropout rates, poor quality, and weak funding, the combination of which contributes greatly to the generation of idle and unskilled youth with few opportunities. Only 53 percent of Dominicans complete primary school and the net secondary school enrolment rate is just 38 percent. What is more, the average score on the grade eight national exam in 2004 was 52 percent (Secretaría de Estado de Educación, 2005). As a result, even the lucky few that stay in the system long enough to graduate tend to leave school with a dearth of relevant skills for a successful school-to-work transition. When compared to other Caribbean countries for which UNESCO data is available, the Dominican Republic's 1.1 percent of GDP in education spending in 2004 was by far the lowest among the Latin American and Caribbean countries with available data.84 The number of youth who can be considered at risk--poor, out of school, and jobless--is therefore large and growing. This increase is believed to be "a key factor in the rise in crime, violence, gangs and other forms of risky behavior, particularly among the poorest segments of this age group, which see fewer prospects for the future" (World Bank, 2005d). 5.23 Violence in Schools. A disquieting number of students in the Caribbean have witnessed physically violent acts in their schools. As a result many students no longer feel safe in their schools and some drop out (Garner et al., 2003). A 2003 representative sample survey of school children in nine Caribbean countries found that one-fifth of the males carried weapons to school in the previous 30 days and one-tenth had been knocked unconscious in a fight. Over 40 percent reported that sometimes or most of the time they think about hurting or killing someone else (Halcon, 2003). Evidence suggests that abuse by teachers of students is common (World Bank, 2003a). In the Dominican Republic, most youth (54 percent) in a recent survey reported that there was violence in the schools, and 42 percent indicated that they knew of violent acts committed on school grounds (ALEPH, 2006). 5.24 Role of Police. The potential role of the police in crime prevention is weakened by problems within the police force. The majority of respondents in youth focus groups in 83Interviews with Dominican National Drug Control Agency 2005. 84Domestic per pupil public expenditure at the secondary level as a share of GDP per capita was just 3.6 percent in 2003, less than a quarter of the LAC average of 17.8 percent. Moreover, it is the only country with available data in the LAC region where public funds for secondary education decreased by more than 20 percent as a share of GDP during the period 1998-2003; over that period the Dominican Republic's funding decreased by 39 percent. World Bank (2006d) 70 the Dominican Republic noted that even if they were caught committing a crime, the possibility of buying one's way out through bribes to police was always a feasible option, as well as that of using padrinos (Godfathers) who could intervene on their behalf and have court decisions suspended, or even have prison inmates freed (ALEPH, 2006). The national police itself reports an average of 36 shooting deaths per month by the members of its force (ALEPH, 2006). In 2005, 18 percent of all violent deaths/homicides were a result of police shootings. Though the ages of the victims are not known, anecdotal evidence suggests that a large number of these are youth: one study cited 23 unprovoked killings of street children by los cirujanos ("the surgeons", a police unit that conducts night sweeps) in three neighborhoods of Santo Domingo over an eight-month period (Commins, 2006). 5.25 Availability of guns and other weapons. The proportion of Caribbean adolescent males who carry firearms is extremely high. The survey in nine Caribbean countries mentioned above revealed that 20 percent of male students had carried a weapon to school in the previous 30 days, nearly as many had been in a fight using weapons (World Bank, 2003a). Evidence suggests that these figures are much higher for those who are not students. In the United States, the equivalent figure for high school students (grades 9-12) was 9 percent, less than half (Centers for Disease Control and Prevention, 2003). In the Dominican Republic, between September 2003 and June 2005 alone, the National Police issued more than 178,000 licenses for owning and carrying firearms, the majority of which were handguns (71 percent), shotguns (14 percent) and pistols (14 percent). According to the National Police, at least 75 percent of homicides are committed using these types of weapons (ALEPH, 2006).85 For more detailed discussion of community-level risk factors and program responses see Annex 5.1. 5.26 Interpersonal level: The most important actors at this level are the family, peer groups, and teachers, each of which can act as protective and risk factors. For the Dominican Republic, available data points to the following concerns at this level. 5.27 Domestic violence, child abuse and corporal punishment. High levels of domestic abuse and corporal punishment throughout the Caribbean are severe risk factors likely to promote future violent behavior. A recent survey revealed that 22 percent of ever-married women aged 15-49 had been victims of physical violence at the hands of a male partner, and 67 percent had suffered emotional abuse (Caceres, 2004). International evidence suggests that children who witness domestic violence are more likely in the future to engage in delinquent and violent behavior (Smith and Thornberry, 1995; Margolin, 1998). Child abuse is also associated with an increased probability that children engage in delinquent and violent behavior, as well as increased risk of children abandoning the home. More than 40 percent of street kids surveyed by Niños del Camino cited abuse in the home as the leading cause of abandoning their families. Corporal punishment continues to be widespread in the Dominican Republic and elsewhere in the Caribbean, in both schools and homes, and particularly against boys (World Bank, 2003a). Much international evidence links use of corporal punishment to later use of 85Data obtained from the National Police Statistics office. 71 violence by children and adolescents; a study in Jamaica has confirmed this link (Meeks, 2001). 5.28 Peers and role models. Most youth in the Caribbean identify parents, entertainers or teachers as role models (Luther, 2002). However, the historical absence of male adult figures in the household for role modeling and mentoring compounds the influence of "negative" role models, particularly for boys. Drug dons are an important source of admiration due to their wealth and power. The drug don and his approachability and interest in recruiting children make him a particularly dangerous role model as youth easily become engaged in his business (World Bank, 2003a). 5.29 Gangsterism and related activities are a large--and growing--problem in the region. According to survey data from the Caribbean Youth Development report, 20 percent of male students and 12 percent of female students surveyed reported having belonged at one point to a gang (World Bank, 2003a). This echoes the marked increase in the number and influence of gangs cited in youth surveys in the Dominican Republic, including one conducted with youth ages 14-17 currently participating in remedial programs where half admitted to belonging to las naciones (ALEPH, 2006). Gangs are highly organized, satisfying the needs of young Dominicans at various levels: at the individual level (through respect, power, authority, recognition, and financial gain), the relationship level (support, caring, friendship, and health services/medical attention), and the community level (rules, training, protection, financial benefits). Gangs are located in all the major cities, though they seem to be expanding into some rural communities. 5.30 Drug trafficking influence. The defining characteristic of Dominican gangsterism is involvement in drug trafficking. The scarcity and low pay of legal jobs; the attraction of the "easy money"; the existence of laws protecting those under 18 from prison (Luther, 2002), and the marketing of drug dons (Barker, 1995) makes involvement in drug trafficking very attractive to youth (World Bank, 2003a). A recent study showed that gang members were 20 times more likely to sell drugs than non-members, and 35 times more likely to collaborate with drug dealers. Even when compared to other vulnerable groups (out of school youth, sex workers, "sankis"86), gang members were still at least three times more likely to be involved in the drug trade. Worldwide, those youth most likely to participate in delinquent or violent activities are usually part of a group, though this association tends to be higher for theft, robbery and rape, and lower for premeditated murder and assault with the intent to inflict grievous bodily harm (United Nations, 2003).87 For a more detailed discussion of interpersonal-level risk factors and program responses, see Annex 5.1. 5.31 Individual level: At the individual level, some of the most influential risk factors for youth violence are biological (being male; delivery complications at birth); psychological/behavioral (degree of self regulation and self esteem; low intelligence and low educational achievement, early sexual initiation); and environmental (exposure to 86"Sankis" are generally male prostitutes that offer services to tourists in exchange for having their entertainment costs covered, usually food and alcohol. 87Available cross country statistical data show that 60-75 percent of all juvenile offenses are committed by members of various groups. 72 violence and conflict in the family; involvement with drugs, alcohol, and tobacco). In the Dominican Republic, some of the key risk factors are the following. 5.32 Drug and alcohol use. Although data on drug and alcohol use are scanty, available evidence suggests a widespread social acceptance of alcohol in nearly all Caribbean countries, and of marijuana in some, among both in-school and out-of-school youth (Barker, 1995). In Jamaica, over three-quarters of students report that alcohol and cigarettes are easily obtainable and 60 percent believe the same about marijuana, while almost half have close friends that smoke "ganja" (National Centre for Youth Development, 2003). In the Dominican Republic, alcohol is very accessible and by far the drug most widely consumed by youth. According to a recent survey conducted among secondary school students, 81 percent had consumed alcohol at some point in their lives, while 85 percent had had the opportunity to consume (Aleph, 2006).88 A separate survey finds the mean age at first consumption of alcohol to be 13 years (Luther, et al., 2002). Two characteristics unique to the Dominican Republic are the high rate of non-marijuana drug consumption (63 percent of drugs reported used were drugs other than marijuana), and the high prevalence of tranquilizers and stimulants consumed, at 12.8 percent (females) and 6.5 percent (males), and 6.2 percent (females) and 4.4 percent (males) respectively. These figures were the highest in a recent OAS report on drug consumption among seven countries in the LAC region (Organization of American States, 2004). 5.33 The age of onset of sexual activity in the Caribbean is the lowest in the world (Blum, 2002, as cited in World Bank, 2003a). Data from the Caribbean Health Survey showed that of the 35 percent of students who reported having had sex, initiation occurred before the age of 13 for nearly two-thirds (including 82 percent of males and 52 percent of females), surpassing all other regions for which data is available (Halcon, 2003). In a related study, early initiation of intercourse was found to be predictive of weapon-related violence and gang involvement (among boys and girls), and alcohol use and running away (among girls) (Ohene, 2005). Teenage pregnancy in particular is a pressing concern as it is an important contributor to delivery complications at birth: the Dominican Republic ranks fifth in the Latin America region in number of births among 15-19 year-olds, nearly double the average for the region (Table 5.2). For more detailed discussion of individual level risk factors and program responses see Annex 5.1. POLICY RECOMMENDATIONS FOR STRENGTHENING THE RESPONSE TO YOUTH VIOLENCE 5.34 The fundamental challenge in assessing the effectiveness of youth violence prevention strategies in the Caribbean--and in the Dominican Republic in particular--is that although a multitude of programs exist, there is virtually no evaluated evidence of such interventions to guide policymakers and practitioners in identifying the relative effectiveness (including cost-effectiveness) of strategies and approaches. The recommendations presented here therefore draw on international research and experiences that have been rigorously evaluated and proven effective in order to: i) identify current programs that merit expansion and highlight elements that could improve 88These figures are for lifetime prevalence (not annual). 73 existing strategies; and ii) suggest complementary strategies to consider, including some which would require a reassessment of current approaches in light of their relative ineffectiveness. 5.35 Table 5.3 presents widely accepted examples of youth violence prevention strategies that have proven effective--and ineffective--in a range of contexts.89 The framework addresses two important considerations: i) the developmental stages of an individual (from early childhood to early adulthood); and ii) the ecological systems through which risks can be addressed (presented in Chapter 3). The strategies presented here are not exhaustive, but are meant to emphasize the spectrum of possible solutions. No single strategy on its own is likely to be sufficient to address youth violence; multiple, concurrent approaches are required. Since much of the current knowledge of effective approaches and programs is based on international evidence, care must be taken to ensure relevance for the particular country and community in which they are implemented. 5.36 In recent years, the Dominican Republic has placed increasing emphasis on finding solutions to the problems of crime and violence in general, and youth violence in particular. As in most countries, this strategy comprises policies that fall into two broad categories: i) those that are control/treatment oriented, using the police, courts and prison system to control the behavior of individuals who engage in violence; and ii) and those that are prevention oriented, either towards the general population or towards individuals considered at risk. Using the international evidence summarized in Table 5.3 and an inventory of over 50 of the most prominent public and non-government programs addressing youth violence in the Dominican Republic,90 this section highlights select strategies and programs that merit particular attention for further evaluation and potential expansion.91 5.37 Provide holistic violence control/prevention approaches targeted to high- violence communities, emphasizing a combination of community policing and improved public services. This type of approach tends to address risk factors at all levels, if appropriately implemented. It targets youth and their families with much needed basic services (early child care, education, health, security) as well as positive alternatives to crime and violence (through second chance programs, extracurricular activities, life skills and occupational training). Although international evidence of the impact of community policing on reducing crimes is mixed, it does improve the public's perception of safety as 89Compiled from WHO (2002) ; WHO (2003) ; World Bank (2006); US Surgeon General (2001); Tolan and Guerra (1994), Sherman (1997) and Kellerman et al. (1997). 90See Annexes 5.4 and 5.5, which are based on Aleph (2006). Programs were inventoried using the ecological framework, assessing both the level of intervention (individual, relationship, community, and society) as well as the level of prevention targeted (primary prevention or universal coverage; secondary prevention targeted to youth at risk; tertiary prevention geared to reducing violence among high-risk youths or preventing further violence from existing perpetrators). 91Programs were selected based on a series of criteria: i) fall within internationally proven strategies for addressing youth violence or have demonstrated effectiveness in the Dominican Republic; ii) have a significant beneficiary population; iii) can be replicated with existing institutional capacity; and iv) where data is available, are shown to be cost-effective. 74 well as the image of the police (WHO, 2003)92 (examples in the Dominican Republic which address part of these objectives include Barrios Seguros and Centros Educativos de Fe y Alegría). 5.38 Expand access/retention in schools for high-risk individuals and communities, with particular emphasis on incentives for early child development and for completing secondary education.93 International research underscores consistently the importance of the protective factor of providing disadvantaged children with a strong start in school through quality early child development programs (ages 0-5), as well as that of keeping children in school--and connected to their school--through their adolescence. Long term follow-up studies of prototypes in a range of countries have shown important effects on reducing violence and other delinquent behavior. Moreover, they have proven more cost-effective in the long run than many other options.94 Examples in the Dominican Republic include the Fund for Early Education (Fondo de Educación Inicial) and the SOLIDARIDAD program. Neither the Fondo de Educacación Inicial (not surprisingly, given its target population) nor the Solidaridad program offers financial support to keep students in secondary school. Given the high rates of school leaving at the secondary level, expanding SOLIDARIDAD from basic to secondary education may be worth considering (see Annex 5.1). 5.39 Improve school quality, relevance and efforts to incorporate violence prevention into the curriculum, and increase the involvement of parents. Together with local NGOs, the Ministry of Education is actively reforming the curricula, enhancing teacher training to improve the quality of primary and secondary education and reduce abuse in schools, increasing the relevancy of education, as well as actively promoting non-violence in the curriculum.95 Through its Decentralization Plan, it also intends to significantly increase parental involvement in school management (World Bank, 2003a). 5.40 Provide "second chances" for youth at risk to complete their formal education, obtain relevant job skills, and/or learn relevant life skills. Programs with flexible schedules designed to allow youth to obtain their primary and secondary education equivalency are a cost-effective way to invest in the human capital development of the large share of youth who have never completed their formal education. Programs offering job skills with job experience (internships) as well as life skills training have been shown to have an important impact on improving the quality of jobs obtained by youth, an important factor in reducing crime and violence (World Bank, 2006b). Examples in the Dominican Republic include Educación Básica Para Adultos 92See, for example, Buvinic and Morrison 2001. See also DESEPAZ Program in Colombia. WHO (2003b) 93Note conditional cash transfer programs in Mexico (Oportunidades), Colombia (Familias en Accion), and Brazil (Bolsa Familia). 94US Surgeon General (2001); Kellerman, A. et al.; World Bank. (2006b); Schweinhart L. et al (2005). 95Secretaría de Educación programs include, among others the Education Strategic Development Plan (2003) Protección Contra la Violencia, Resolución de Conflictos, Plan de Descentralización, Proyecto de Apoyo para la Calidad Educativa, Uso Indebido de Drogas, Proyecto Multigrado Renovado (see Annex 5.4 for further details) 75 Table 5.3: Violence prevention strategies by developmental stage and ecological context Early Childhood Middle Childhood Adolescence Early adulthood (0-5 years) (6-11 years) (12-19 years) (20-29 years) *Providing incentives for youth at high risk for violence to complete secondary schooling (a) *Providing incentives to pursue courses in higher * Academic enrichment programs for those at risk of education *Early Child Development and *Parenting training (a) dropping out Preschool enrichment programs for at *Individual Counseling (b) * Parenting Training Individual risk children (a) *Programs providing information *Probation or parole programs that include meetings with * Preventing unintended pregnancies about drug abuse (b) prison inmates describing the brutality of prison life (b) * Job Training combined with life skills and * Parenting training *Residential programs in psychiatric or correctional internships * Increasing access to prenatal and institutions (b) postnatal care *Programs providing information about drug abuse (b) * Residential programs for offenders (b) * Training in the safe use of guns (b) *Programs modeled on basic military training (b) * Psychotherapy for high-risk youth and offenders (b) *Trying young offenders in adult courts (b) *Mentoring Programs (a) *Mentoring programs (a) *Home visitation (a) * Targeted incentives to mother to *Family therapy (a) * Programs to strengthen ties to family and jobs, and Relationship keep child in school (a) *Temporary foster care programs for serious and chronic reduce involvement in violent behavior (e.g. family, *Training in parenting (a) * Home-School partnership delinquents programs to promote parental *Peer mediation or peer counseling (b) peers) * Marital and family therapy involvement * Gang membership prevention/intervention programs (b) *Creating safe routes for youths on their way to and from *Establishing adult recreational programs *Creating safe routes for children school or other community activities *Community policing on their way to and from school or *Improving school settings, including teacher practices, * Proactive arrests *Monitoring lead levels and removing other community activities school policies and security * Directed patrols toxins from homes *Extracurricular activities/supervised after-school programs *Reducing availability of alcohol *Improving school settings, * Positive youth development programs *Improving emergency response, trauma care and * Increasing the availability and quality including teacher practices, school * Supporting classroom management techniques access to health services of early child development and child- policies and security * Behavior modification via " thinking" skills * High quality rehabilitation programs Community care facilities * Providing "second chance" education programs * Imprisonment of high-rate career criminals *Providing after-school programs (equivalency) * Prison-based drug rehab programs to extend adult supervision * Life skills training * National Youth Service Programs *Training health care workers to identify and refer youths at * Corporal punishment in schools (b) * Extracurricular activities high risk for violence * Citizen patrols (b) *Community policing * Social casework: counseling, close supervision and *Reducing the availability of alcohol social services (b) *Improving emergency response, trauma care and access to * Reactive arrests (b) health services * Random patrols (b) * National Youth Service Programs *Buying back guns (b) *Buying back guns (b) * Trying young offenders in adult courts (b) *Deconcentrating poverty *Deconcentrating poverty *Deconcentrating poverty *Deconcentrating poverty *Reducing income inequality * Reducing income inequality *Reducing inequality *Reducing income inequality *Job creation programs for the chronically unemployed Societal * Reducing media violence *Reducing media violence *Public information campaigns *Public information campaigns * Public information campaigns *Public information campaigns *Reducing media violence *Promoting safe and secure storage of firearms *Reforming educational systems *Enforcing laws prohibiting illegal transfers of guns to youth *Strengthening and improving police and juvenile *Strengthening and improving police and judicial systems justice systems *Reforming educational systems Sources: Compiled from WHO (2002); WHO (2003); Blueprints (2003); World Bank (2006b); U.S. Department of Health and Human Services (2001); Tolan and Guerra (1994), Sherman (1997) and Kellerman et al. (1997). (a) Demonstrated to be effective in reducing youth violence or risk factors for youth violence (b) Least promising or shown to be ineffective in reducing youth violence or risk factors for youth violence 76 and PREPARA (SEE), Juventud y Empleo (SET) and the Aprendices con Don Bosco after-school program. 5.41 Promote strategies using existing youth serving organizations to increase capacity for home visitation and parental training to reduce levels of violence in the home. Home visitation is a proven approach to reducing youth violence and delinquency which exists in many parts of the world. It is targeted to low-income young mothers and families who are expecting or have recently had their first child, and those at increased risk of abusing their children with the objective of providing parenting training, support, counseling, child development monitoring and referrals to outside agencies. It has proven effectiveness in a range of countries with significant long term effects in reducing violence and delinquency, as well as other risk factors such as alcohol abuse (U.S. Department of Health and Human Services, 2001; WHO, 2003; Olds DL et al, 1998; Farrington and Welsh, 1999; Blueprints, 2003.) Examples in the Dominican Republic include PROFAMILIA, IDDI, Don Bosco, Progresando (see Annex 5.5). Complementary Strategies to Consider 5.42 Provide positive role models through nation-wide mentoring programs. Well- conducted mentoring programs, which tend to be relatively cost-effective, can reduce aggressive behavior, delay the onset of drug and alcohol usage among youth and boost school retention and performance. These programs also have significant impacts on reduced substance abuse, reduced skipping school, and improved relationships with their families (World Bank, 2003a; U.S. Department of Health and Human Services, 2001; Tierney et al, 2000). 5.43 Open schools after hours and on weekends for supervised extra-curricular activities, training and conflict resolution in high-crime communities. Peak times for juvenile crime are during the hours immediately after school, yet many youth are unsupervised after school because their parents work. After-school programs hold great potential value and deserve serious consideration in community prevention planning.96 According to a UNESCO study, participating schools in a similar program in Brazil demonstrated as much as a 60 percent reduction in violence, as compared to other schools in the area. The program also reduced the rates of sexual aggression, suicide, substance abuse, theft, and armed robbery (World Bank, 2006b). 5.44 Pilot national youth service programs, or service-learning programs which give youth practical work experience and life skills while simultaneously helping to meet key development objectives at a national or community level. Jamaica has a National Youth Service program that targets youth 17-24 who are out of school and unemployed. It combines life skills training with 8 months of work exposure in jobs such as teaching aids, health facilitators, early childhood caregivers, environmental aides and information technology (IADB, 2002). Controlled longitudinal studies of the American Conservation and Youth Corps aimed primarily at disadvantaged youth 16-24 found that 96Several controlled studies have found that well-supervised after-school programs substantially reduce juvenile crime, drug use and vandalism. For more information see: Inter-American Development Bank (2002) and Sickmund et. al (1997). 77 program participants in a 15-month follow up were much more likely to have worked for pay and were less likely to be arrested vis-à-vis control groups (Jastrab. et al, 1996, 2004; World Bank, 2005c). 5.45 Several strategies have proven effective in reducing significantly the rates of rearrest and recidivism for youth with a history of chronic or violent criminal behavior (Lipsey and Wilson, 1998; U.S. Department of Health and Human Services, 2001). Two model programs for these youth use a family therapy approach, and place strong emphasis on skills training and behavioral change for youth and their parents. These include: i) functional family therapy; and ii) multidimensional treatment foster care using a clinical intervention targeting individual youth with severe criminal behavior and placing them in trained foster home care as an alternative to incarceration, group or residential treatment. Meta analyses have shown that community-based treatment is more successful than residential treatment and that both types of programs have significantly reduced the number of days of incarceration, overall arrest rates, drug use, and program dropouts versus control groups (U.S. Department of Health and Human Services, 2001). 5.46 Reduce emphasis on "mano dura" programs that emphasize harsher penalties, increased arrests, and more police controls. Criminalization, incarceration and suppression are inefficient responses to chronic youth offenders, gang members, and those committing violent crimes. Despite being a common official response by governments, these "mano dura" programs have proven to be the least effective among a range of policy options (World Bank, 2006b). One promising justice system approach with strong deterrent effects is an intensive protective supervision strategy which removes delinquent youths (status offenders) from criminal justice institutions and provides them with proactive and extensive community supervision with trained professionals (U.S. Department of Health and Human Services, 2001). 5.47 Apply existing laws separating incarcerated youth and adults. The imprisonment of child offenders with adults (due to the lack of facilities for young delinquents) is common in the Caribbean (Singh, 2001, as quoted in World Bank, 2003a). The principle of incarcerating young offenders separately from adults has been accepted in the Dominican Republic, but its application is minimal, with many youth ending up in state prisons. Prison conditions range from "poor" to "harsh", with overcrowding being a pressing concern (U.S. State Department, 2006).97 Improving prison conditions and separating young offenders are important remedial policies as both concerns are associated with higher recidivism.98 5.48 Enforce new legislation requiring the registration of guns. Currently, it is enforced by occasional checkpoints where vehicles must produce the certificates of registration for any firearms in the car. Banning guns during periods known for high violence/homicide could also be considered, such as weekends and holidays. This 97In 2005, 145 people died in a fire in a cell block that was designed to hold 80. 98 There are 4 prisons for minors separated by sex. Minors can be prosecuted in either corrective or criminal status. Corrective status is for minor crimes (simple robbery, delinquency, drug use) whereas criminal status is for larger crimes (murders, aggravated robbery). Corrective imprisonment ranges from a few months up until 3 years whereas criminal imprisonment ranges from 1-10 years. 78 approach was successfully implemented by the municipalities of Cali and Bogota, which witnessed lower homicide rates as high as 20 percent when the ban was in effect (Guerra, 2006; World Bank, 2003b). 5.49 Restrict the availability of alcohol. This has demonstrated positive effects on reducing violence (both criminal and domestic) as well as criminal offenses (both serious and minor related to property and traffic) (WHO, 2003). Given the widespread acceptance of alcohol consumption even at very early ages in the Dominican Republic, a concerted effort to treat alcohol as a drug in all anti-drug programs may have considerable impact on reducing violence. Policies such as higher taxation of alcohol combined with greater enforcement of minimum legal age have demonstrated success in reducing consumption among young people in some countries (WHO, 2003b). In Colombia, alcohol sales were restricted with closing times imposed on bars and nightclubs (Guerrero and Concha-Eastman, 2001). CONCLUSIONS 5.50 The most important conclusion of this chapter is that youth violence is a legitimate concern for the region, but it is not an intractable problem. Youth violence is preventable. Although deaths and injuries from youth violence constitute a major public health, social and economic problem across the Caribbean, including the Dominican Republic, a broad range of viable strategies for preventing and reducing youth violence exists. Moreover, it is important to remember that youth are not "the problem," but rather a product of the individual, family, community and social environments in which they live. 5.51 In addition to the thematic conclusions presented above, several cross-cutting recommendations emerge. First, there is an urgent need to develop and strengthen data systems which regularly compile and monitor trends in youth violence (including injuries and deaths, both as victims and perpetrators) across key institutions. Current data is extremely weak and inconsistent, and there is little collaboration across entities. In the Dominican Republic, the Procuraduría General could establish a system by which routine information available from health services, emergency departments, the police, the criminal justice system, and other authorities relevant to youth violence can be compared on an annual basis. Strengthening these systems will provide valuable information for formulating policies and for evaluating them.99 5.52 Second, there is a clear need to generate scientific evidence on the patterns and causes of youth violence in specific social settings, as well as the cost to society of such violence. Perhaps more important, however, is the need to generate knowledge on what works to reduce violence through rigorous evaluation of both impact and costs. Only with consistent standards of systematic evaluation of interventions can policy tradeoffs be 99Uniform standards for defining and measuring youth violence should be incorporated in regular surveillance systems. Included in this should be methods to establish the ratio of fatal to non-fatal cases of violence-related injuries, classified by the method of attack, age and sex of the victim. Such data can then be used to estimate the magnitude of the youth violence problem where only one type of data ­ such as mortality or morbidity ­ is available. See WHO (2003). 79 legitimately addressed.100 In particular, there is a need for: i) longitudinal studies evaluating long term impacts of interventions conducted in childhood; ii) evaluations of the impact of interventions in sectors sometimes considered unrelated (e.g., education, health, employment, etc.) on reducing youth violence and other risk factors; iii) studies on cost-effectiveness of prevention (and control) programs. 5.53 Finally, there is a need to improve public awareness of the availability of ongoing interventions, with particular attention paid to high-risk communities and youth. A public information campaign should not be limited to potential beneficiaries, but should also extend to public and civil society groups working in this area as there is considerable lack of knowledge of existing programs in other institutions. Indeed, in the Dominican Republic, as well as across the sub-region, there are a multitude of programs available to address many of the risk factors facing youth. Evaluating their effectiveness and sharing this knowledge across countries with similar issues to confront will contribute enormously to future success in reducing youth violence. 100Consistent standards include: i) the application of experimental design; ii) evidence of a statistically significant reduction in the incidence of violent behavior or violence-related injuries; iii) replication across different sites and different cultural contexts; and iv) evidence that the impact is sustained over time. 80 6. CASE STUDY: CRIMINAL DEPORTATIONS AND JAMAICA It has been hypothesized that recent crime troubles in the Caribbean could be tied to the activities of deportees who have learned criminal behavior in the developed countries. This report examines the situation in Jamaica, which is one of the Caribbean countries proportionately most affected by criminal deportations. With currently available data, it is possible to conclude that it is unlikely that the average deportee is committing crime in Jamaica. At the same time, it is possible that a minority of deportees is involved in criminal activity. Assisting in reintegration efforts for deported offenders could be a cost- effective way for deportee-sending countries to promote development and weaken international crime networks. 6.1. Each year, the United States, the United Kingdom, and Canada deport thousands of people convicted of various crimes to their countries of citizenship in the Caribbean. Between 1998 and 2004, the United States alone deported over 31,000 convicted criminals to the Caribbean. This chapter focuses exclusively on the issue of criminal deportees: people who have been forcibly returned to their country of origin due to their conviction for a criminal offense, such as drug offenses, violent crime, and immigration- related crime. The chapter does not discuss the less controversial practice of deporting those who have entered the country without a visa or violated the terms of their visa or residency permit.101 6.2. There are few topics in Caribbean criminal justice more contentious than criminal deportation. On the one hand, as the international community has recognized, "It is the right of every nation State to decide who can enter and stay in its territory and under what conditions" (United Nations, 2004). Residence permits are a privilege granted to non-citizens contingent on their good behavior. Clearly, the commission of a crime does not constitute "good behavior," and few would deny the right of host countries to revoke the permits of convicted criminals. This practice is widespread, and Caribbean countries themselves deport non-citizen convicts on a regular basis. 6.3. On the other hand, there is a widely held belief in the Caribbean (as well as in Central America) that recent crime troubles can be tied directly to the activities of deportees "schooled" in crime in the developed countries, especially the United States, Canada, and the United Kingdom.102 This chapter looks at the situation in Jamaica, which is proportionately most affected by criminal deportations, and where officials have worried that the country's rising murder rate may be linked to the growing stock of 101Criminal deportations are distinct from deportations for simple immigration violations like entering the country without a visa or permit. However, the ranks of criminal deportees do include those deported for crimes that are related to immigration. Examples of immigration-related crimes include modifying visa papers, attempting to bribe an immigration official, lying on an immigration-related document, and smuggling migrants. 102This view is widely held enough that the CARICOM representatives recently suggested that the Inter- American Commission on Human Rights intervene to reduce the flow of deportees. See Smith, S, "Carib states want a study on the link between deportation and crime." Jamaica Gleaner, 16 December 2005. 81 deported convicts.103 Figure 6.1 shows the figures for both homicides and deportee arrivals in recent years. No clear relationship emerges from this graph. Figure 6.1: Number of Deportees and Number of Murders in Jamaica 8000 14711500 onsitatr 7000 1139 6000 1045 953 5000 1000 sredru poeDl 975 4000 849 887 3000 nai 500 Mforeb 2000 mirC m 1000 Nu 0 0 1998 1999 2000 2001 2002 2003 2004 USA deportees (left axis) UK deportees (left axis) Canada deportees (left axis) Murders (right axis) Source: Social and Economic Survey of Jamaica; Jamaica Constabulary Force. 6.4. Some Caribbean commentators have argued that it was conditions experienced in the developed countries, not in their countries of citizenship, which drove these people into criminality. For example, Dr. Prem Misir, Pro-Chancellor of the University of Guyana, asserts, "...criminal deportees have been intensively socialized in the criminal fields in the U.S. These deportees are in full possession of their U.S. criminal tool kit."104 If Dr. Misir is correct, the question then becomes: who is responsible for criminals born in the Caribbean but made in the developed world? And who should bear the burden of this population, the large and wealthy countries or the small Caribbean countries? 6.5. Of course, the issues are not as clear cut as this. Deportees are not a homogenous group. There are cases of deportees who, in all but their paperwork, were citizens of the developed countries, having been raised there and holding few connections to the countries of their birth. But there are also those who were career offenders in their home countries before venturing out into trans-national organized crime, and those who fall somewhere in the middle, whose criminality may be attributed to experiences in both countries. And even if it were possible to determine where a criminal is made a criminal, this would not answer the broader policy question: how can developing countries cope 103In his address at the Opening of the Twenty-Fifth Meeting of the Conference of Heads of Government of the Caribbean Community on 4 July 2004, in St. George's, Grenada, then Jamaican Prime Minister PJ Patterson referred to "security concerns arising from the nexus of the narcotics trade, gun-smuggling and criminal deportees." He used similar language in a speech at the CARICOM 30th Anniversary Lecture Series at Medgar Evers College in New York on 2 October 2003, where he referred to "the U.S.-based nexus of the narcotics trade, gun smuggling and criminal deportees." In his address to the Joint Session of The St. Lucian Parliament on 1 July 2004, he referred to "the impact of deportees on the escalating rate of crime." 104Misir, P., "Government tackling globally-rooted crimes." Accessed on the website of the Guyana Government Information Agency: http://www.gina.gov.gy/archive/researchp/rpgovtacklingcrimes.html 82 with regular injections of uprooted convicts? And how can this problem be resolved to reduce the spread of transnational criminality? 6.6. This chapter lays out the latest available data on the scale and nature of criminal deportation from the United States, the United Kingdom, and Canada to Jamaica and suggests some ways the issue might be resolved for the benefit of all concerned. THE SCALE OF THE EXPATRIATE POPULATION 6.7. To appreciate the scale of the deportations, it is necessary to understand the size of the Caribbean expatriate community. An estimated three million Caribbean-born people were living in the United States at the time of the 2000 Census, about 10 percent of the total immigrant population in the U.S. and about 1 percent of the U.S. national population. The 300,000 Caribbean-born people living in Canada at the time of its 2001 Census also make up about 1 percent of the population of that country. In the United Kingdom, some 250,000 people born in the Caribbean were counted during the 2001 Census. 6.8. In total, 3.5 million Caribbean-born people were living in these three countries in 2000/2001. The figure is perhaps 3.8 million Caribbean-born people as of 2006,105 more than are presently living in the countries of Montserrat, Anguilla, British Virgin Islands, the Turks and Caicos Islands, the Cayman Islands, St. Kitts and Nevis, Antigua and Barbuda, Dominica, Grenada, Aruba, St. Vincent and the Grenadines, St. Lucia, the Netherlands Antilles, Barbados, the Bahamas, Martinique, Guadeloupe, Suriname, and Guyana combined. 6.9. About 85 percent of these migrants live in the United States, with the largest pool from Cuba, followed by the Dominican Republic, Jamaica, and Haiti (the four largest independent countries in the Caribbean). Figure 6.2 shows the Caribbean-born populations in the U.S. from several countries. Jamaica has the largest share of its population living in the United States. For every five Jamaicans living at home in Jamaica, there is one living in the United States, with lower ratios for Cuba (9 percent), the Dominican Republic (8 percent), and Haiti (5 percent). According to U.S. Census projections, the U.S. Jamaican-born population was over 600,000 in 2005 (Department of Homeland Security, 2005). 6.10. In Canada, Jamaicans are the most populous group of Caribbean origin living in the country, with just under 122,000 Jamaican-born residents, nearly twice as many as the next largest Caribbean group (Trinidadians), according to the 2001 Census. About 150,000 of the 250,000 Caribbean-born people living in the United Kingdom were Jamaican, according to the 2001 Census. In total, about 683,000 Jamaicans were living in the three countries in 2000/2001, or more than a quarter of the total living in Jamaica at that time. Because these estimates are based on census figures, which may undercount undocumented immigrants, the true number of Jamaican expatriates may be higher. 105U.S. census projections suggest a Caribbean population of 3.211 million in 2005. Similar projections are not available for Canada and the United Kingdom, but if similar growth were experienced, some 600,000 Caribbean nationals might be seen between the two countries. 83 Figure 6.2: Caribbean-Born Populations in the United States 1,000 952 900 800 737 nds 692 700 608 ousahTni 600 500 439 385 411 400 348 334 ontilau 300 225 197 200 169 Pop 79 92 100 61 69 12 528 25 0 Cuba Dominican Rep. Haiti Jamaica 1960 1970 1980 1990 2000 Source: Lapham et al., 1993. CRIMINAL DEPORTATION 6.11. Many migrants arrive with limited resources, and some arrive outside the law. Poor migrants may be forced to settle in some of the most crime-ridden neighborhoods of their new countries, and may face ongoing marginalization. It is not surprising that some engage in criminal behavior, are arrested and imprisoned. Whether migrants are more likely to commit crime is hotly debated, and resolving this issue is hampered by a lack of clarity on the number of undocumented migrants (Rumbaut et al., 2006).106 6.12. By mid-2004, federal and state prisons in the United States held over 90,000 non-citizens, representing 6.5 percent of the overall prison population and over 20 percent of the federal prison population. The burden on state prisons varied considerably between regions of the country. In the high-immigration state California, over 10 percent of prisoners were non-citizens (Harrison and Beck, 2005). Some 29 percent of federal drug defendants convicted during 2003 were identified as non-citizens, and more than half (56 percent) had at least one prior adult conviction (Bureau of Justice Statistics, 2005). 6.13. Concerns about immigrant crime and the demands it was placing on the U.S. criminal justice system contributed to the passage of the 1996 Illegal Immigration Reform and Immigrant Responsibility Act. This Act made significant changes to the deportations regime, reducing appeals and greatly expanding the definition of deportable "aggravated felonies" to include a range of lesser offenses.107 Resources were also 106At least one study has found that first generation immigrants are less likely to be convicted of a crime than the native born. 107The term of "aggravated felony" first appeared in the immigration context in 1988 in the Anti-Drug Abuse Act, where it was limited to murder, drug trafficking and firearms trafficking. This was expanded by subsequent legislation and related case law to include a much wider range of offenses, including most violent crime, theft, and immigration-related offenses, such as document fraud and perjury. The 1996 Act amends the definition of "aggravated felony" by, among other things, lowering the fine and sentencing 84 directed to ensure that every deportable convict was, in fact, deported. As a result, the number of criminal deportees from the U.S. to the Caribbean more than doubled between 1994 and 2004 (see Figure 6.3). Figure 6.3: Total Criminal Deportations from the U.S. to the Caribbean 6,000 5,166 5,000 4,607 4,799 4,574 5,018 3,914 4,426 4,000 4,342 3,705 3,000 2,499 3,135 2,698 2,000 2,290 1,000 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Source: U.S. Department of Homeland Security.108 6.14. If the number of prisoners in U.S. federal prisons is examined as a share of the total populations of the home countries of the convicts, then Jamaica contributes the most prisoners per head of population. Not surprisingly, then, Jamaica is the country in the Caribbean with the largest flow of deportees relative to its population, and its lead is growing (see Figure 6.4). An average of 1,200 convicts per year were deported from the U.S. to Jamaica between 1993 and 2004. 6.15. Jamaicans are also the most deported Caribbean population group from the United Kingdom and Canada, largely because they are by far the largest Caribbean population in those countries. Between 2001 and 2004, Jamaica absorbed an average of 2,700 convicts a year from the three countries. In 2003, Jamaica's own current prison population was 4,744 (Walmsley, 2005), so the influx was equivalent to releasing more than half the domestic prisoner population into society every year. In one study on released prisoners in the United States in the mid-1990s, about two-thirds were re- arrested within three years.109 It is therefore not surprising that Jamaicans might be concerned about an inflow of convicts of this magnitude. thresholds for many offenses, effectively including relatively minor crimes. Under this law, criminal deportation has been ordered for crimes such as shoplifting and urinating in public. See U.S. Department of Justice Fact Sheet 03/24/97. 108Data published (Table 43) online at: http://www.uscis.gov/graphics/shared/aboutus/statistics/ENF03yrbk/ENF2003list.htm 109Data from the United States Bureau of Justice Statistics, accessed on: http://www.ojp.usdoj.gov/bjs/reentry/recidivism.htm 85 Figure 6.4: Criminal Deportees from the U.S. per 100,000 Population of Home Country 70 61 60 52 50 40 36 32 29 30 27 21 20 16 10 5 0.6 0.5 3 0 Cuba Haiti T&T Dominican Bahamas Jamaica Rep. 2000 2004 Source: U.S. Department of Homeland Security, 2005. Figure 6.5: Total Criminal Deportations to Jamaica: 1998-2004 2500 2000 3 7 71 1982 1993 1871 16 153 154 1500 1483 1410 1462 1274 1000 765 5 500 0 2 3 25 32 7 3 2 24 22 24 150 20 141 167 20 0 1998 1999 2000 2001 2002 2003 2004 USA UK Canada Source: Social and Economic Survey of Jamaica, various years. 6.16. During the late 1990s, the United States was far and away the leading source of criminal deportations to Jamaica, but then deportations from the United Kingdom began to rise rapidly. While the bulk of these deportations were due to immigration-related crime, the number and the share of drug-related deportations increased dramatically from 2001, nearly reaching U.S. levels by 2004. Proportionate to the expatriate Jamaican population in each country, the U.K. today has a greater rate of deporting drug offenders than the U.S. This may be due to the fact that cocaine markets are growing in the U.K., 86 and Jamaican "yardie" groups have traditionally been associated with marketing this drug (see National Criminal Intelligence Service, 2005).110 6.17. Despite this shift, the U.S. has continued to expel by far the largest number of violent offenders, deporting over 200 convicted murderers and 128 sex offenders to Jamaica between 2001 and 2004, while the U.K. and Canada combined deported 24 murderers and 32 sex offenders between them. On the whole, however, 81 percent of the criminal deportees sent from the three countries between 2001 and 2004 were deported for immigration-related matters, fraud, and drug offenses, which include an unspecified share of drug possession cases111 (see Annex 6.1 for detailed statistics on deportations to Jamaica from Canada, the U.K., and the U.S. for the 2001-2004, period, disaggregated by reason for deportation). Figure 6.6: Drug Deportations to Jamaica from the United States and the United Kingdom 1200 984 923 1000 911 773 800 864 600 629 400 200 294 0 159 2001 2002 2003 2004 UK USA Source: Social and Economic Survey of Jamaica, various years. 110According to National Criminal Intelligence Service of the United Kingdom, "Hitherto, criminal groups of West Indian origin, mostly Jamaican, were seen to be most prominent in distributing crack cocaine within the U.K. ... However, as [cocaine use] has grown, there have been opportunities for others to become involved... [Today,] most of the detected movements of cocaine trafficked from the Caribbean to the U.K. are smuggled by couriers, many under the control of Jamaican and Trinidadian groups." 111Headley's research, cited below, found that about half those Jamaicans deported for drug offenses between 1997 and 2003 were convicted of drug sales and half for drug possession. 87 6.18. Thus, the vast majority of the offenders deported (81 percent) were convicted of non-violent crimes. Of course, those deported on the basis of non-violent offenses could be violent criminals, and drug dealers in particular may be more likely than other expatriates to commit violence. It is often easier for police to deport suspected gang members for immigration violations or drug possession, for example, than to bring them to trial for other offenses. Figure 6.7: Breakdown of Criminal Deportees by Crime Type, All Three Source Countries, 2001-2004 Other 4% 48 Drugs 38% Fraud and 47 immigration 5 43% 136 49 Firearms and Robbery, violent crime burglary, 11% larceny 4% Source: Elaborated from Social and Economic Survey of Jamaica, various years. 6.19. The Ministry of National Security and Justice study cited by the CARICOM Regional Task Force report (2002) found that a majority of deportees had been away for more than five years and that 14 percent of the 1,730 persons deported in 2001 had been away for more than 20 years. The report suggests that deportees who have been away for a long period of time are at higher risk for recidivism because they are likely to be without a social support system in Jamaica. 6.20. Some commentators have argued that many deportees left their home countries at a young age and learned criminal behavior while abroad. To examine this hypothesis empirically, Bernard Headley, a Professor of Criminology at the University of the West Indies, examined 5,174 records of criminals deported from the U.S. between 1997 and 2003. He found that the mean age of entry to the United States was 23 years, nearly out of the high risk age demographic of 15-25, and less than 3 percent arrived before the age of five years (Headley et al., 2005). A fifth arrived during the formative period of 16-20, however, during which experiences in both countries could have had some effect on the course their lives would later take. Thus, while there are some cases of deportees who left Jamaica at young ages and were largely raised in the United States, they constitute a minority of total deportees. 88 Figure 6.8: Age on Arrival in the U.S. of Jamaican Criminal Deportees (Percent of Headley's Sample) 25 21 20 16 15 t 15 14 cenrep 11 10 9 6 5 5 3 0 Less 6-10 11-15 16-20 21-25 26-30 31-35 36-40 41 and than 5 older years Source: Headley et al, 2005. 6.21. Headley also examined the age at deportation, finding that the vast majority of convicted deportees were no longer young on their arrival in Jamaica, with 62 percent being 31 years or older. Normally, these older men would be less likely to re-offend, particularly with regard to the sort of gang violence with which the deportees are often thought to be associated. 6.22. In summary, the data indicate that the average age of a criminal deportee entering the United States was 23, the average age on deportation was 35, and 81 percent were returned to Jamaica for non-violent offenses. While there is clearly heterogeneity in the pool of deportees, the average deportee does not fit the profile of an individual who is likely to be a violent criminal on return to Jamaica. 6.23. On the other hand, Headley's research also shows that over half (51.2 percent) of the deportees had already been convicted of a crime in the United States prior to the one for which they were deported. And even if the majority of the criminal deportees were deported for non-violent offenses, 224 convicted murderers were included in the flow between 2001 and 2004. Relative to the Jamaica's population, this is not a particularly small number. 89 Figure 6.9: Age on Deportation from the U.S. of Jamaican Criminal Deportees (Percent of Headley's Sample) 25 23 20 20 19 19 16 15 tnecr pe 10 5 3 0 20 or younger 21-25 26-30 31-35 36-40 41 and older Source: Headley et al, 2005. 6.24. With current available data, it is possible to conclude that it is unlikely that the average deportee is committing violent crime in Jamaica. At the same time, it is possible that a minority of deportees is involved in criminal activity, and a few anecdotal cases have been reported in the press. In such small countries, it does not take a large number of offenders to have a large impact, particularly if they assume a leadership role in criminal gangs on their return or provide perverse role models for youth. Specifically with regard to drug trafficking, their transnational connections and criminal experience could make criminal deportees well-suited for this role. 6.25. Are deportees contributing significantly to Jamaica's crime rate? The clearest way of answering this question would be to keep tabs on the deportees received and determine what share are later charged with an offense. Something like this has actually been done in at least two Caribbean countries: Barbados and Trinidad and Tobago. According to Griffin (2002), of 332 criminal deportees returned to Barbados between 1994 and 2000, only 43 (13 percent) had been charged with a criminal offense at the time of the study. The average length of time between arrival and being charged with a crime was 17 months, and, with the exception of one murder case, most of the deportees were charged with burglary or drug-related offenses. Similarly, in Trinidad and Tobago, of the 565 deportees received between 1999 and 2001 only 83 (15 percent) had been charged with a crime, of whom almost half (47 percent) were charged with larceny or drug offenses (Griffin, 2002). 6.26. The CARICOM Regional Task Force on Crime and Security (2002) reviews the same figures and points out that these figures suggest that crime rates among deportees in those two countries are lower than those for the local criminal population. The Task 90 Force argues that the reoffense rate for local criminals in both Barbados and Trinidad and Tobago, and in Jamaica as well, is greater than 50 percent. 6.27. If data is not available that allows us to follow deportees after their arrival in Jamaica, it should still be possible to determine whether deportees are more or less likely to be convicted of a crime than the local population by examining the share of the prison population that has experienced deportation. Of course, Jamaica's low conviction rates (about 25 percent) mean that a lot of suspected criminals are never convicted, so the prison population may not be a true reflection of the criminal population of the country. POLICY IMPLICATIONS 6.28. Whether the deportees are responsible for rising crime in Jamaica is a researchable question. Much support could be brought to this contention if the share of those charged with or convicted of crime were shown to disproportionately contain deportees. The government of Jamaica is currently sponsoring research in which deportees, gathered through responses to newspaper ads soliciting interviews, are questioned about their pasts and their behavior.112 Whatever the conclusions of this research, the question of who is responsible for these people will remain. 6.29. The United States, the United Kingdom, and Canada absorb much of Jamaica's talent, importing nurses, teachers, athletes, and skilled professionals. In fact, an estimated 85 percent of Jamaica's skilled labor emigrates, largely to these three countries (Ozden and Schiff, 2006). Legally, sovereign states are privileged to eject those parts of the migrant population that do not meet standards. Morally, though, these countries might feel some sense of responsibility for the deportees, especially for those who were raised in the ghettos of three of the richest countries in the world. 6.30. There are several possible approaches these powerful countries could take to cushion the blow of criminal deportations on the islands to the south. In the United States, for example, most convicts are released from prison into some form of community supervision. Conditions are also placed on their release, encouraging legal employment and discouraging association with former criminal associates, for example. These protections and conditions are not present when a convict is deported. 6.31. Of course, providing parole officers and halfway houses involves some costs, and a key reason for the current drive in immigration enforcement, at least in the U.S., is to reduce the country's massive corrections bill. But these costs would likely be less if the programs were conducted in Jamaica. For the convicts that left Jamaica as children, whose families may remain abroad, some form of cultural orientation and networking would be essential if a return to criminality is to be avoided. 112A forthcoming study conducted by Jamaica's Ministry of National Security and the Planning Institute of Jamaica, will examine the relationship between deportees and crime. The study was not available for review at the time of this writing. CARICOM is also carrying out a survey on deportees in some member countries, the results of which are expected to be available in early 2007. 91 6.32. The CARICOM Regional Task Force on Crime and Security (2002) recommended that member countries establish Offices for the Resettlement of Deportees modeled after a program in St. Kitts and Nevis, where the Returning Nationals Secretariat is charged with facilitating reintegration of deportees. The Secretariat provides counseling and offers assistance in finding jobs, locating housing, and using social services. In Jamaica, a church-based group known as the Land of My Birth Association has recently started to offer similar services to some deportees (Davidson, 2006). 6.33. The industrialized countries provide a variety of forms of aid to the Caribbean to support development. Subsidizing reintegration for deported offenders could be a very cost effective way of achieving such an end. It would save society the cost, as well as the trauma, of recidivism; reduce criminal justice costs involved in processing and incarcerating repeat offenders; and promote the stability essential to attracting investment and promoting tourism. The end result might be less emigration, reducing the burden of deportation for everyone. 6.34. Moral obligations aside, it is in the self-interest of the United States, the United Kingdom, and Canada to avoid returning dangerous convicts to environments where they are highly likely to offend again. The United States in particular does not benefit from having unstable states just outside its borders. Drug dealers are likely to make use of their connections in both countries to promote further trafficking, and all three countries suffer from cocaine and cannabis imported via the Caribbean or by Caribbean nationals. Exporting criminals could contribute to the building of transnational criminal networks. 6.35. Deportation thinking seems to rest on the premise that the borders are impregnable, when anyone involved in immigration enforcement knows that this is far from the case. Few countries would consider allowing early release of convicts in order to "deport" them to a neighboring city, when this is essentially what deportation accomplishes. It provides immediate relief from corrections expenses, but it releases offenders into an increasingly mobile global community. Headley's sample included over 500 cases where the subject was being deported for at least the second time in the six years included in the study. 6.36. In short, it is not possible for these three countries to simply export their crime problems south. Coming to terms with transnational crime will require dealing with problematic migrants cooperatively. 92 7. CASE STUDY: DRUG TRAFFICKING AND THE NETHERLANDS ANTILLES The drug trade is a prime driver of crime across the Caribbean. In the Netherlands Antilles, authorities estimate that 75 percent of crime is drug-related. Some 60 percent of all the cocaine seized in the Caribbean in 2004 was seized in the Netherlands Antilles, and cocaine seizures there increased dramatically between 2001 and 2004. Confronted with large numbers of people attempting to smuggle drugs by commercial flights, authorities implemented a "100% Control" policy of screening large numbers of passengers for drugs. Drugs were confiscated from the couriers, but in most cases the couriers themselves were not arrested. The program has been very successful in reducing cocaine smuggling via air courier, and could be tested in other contexts, including other Caribbean countries suffering from drug transshipment. The Antillean example also highlights the need for cooperation between Caribbean transshipment countries and destination countries in maritime interdiction. 7.1. Despite their diversity, the Caribbean countries share a common affliction: they are geographically positioned in the world's largest drug transit zone. South America produces nearly all the world's cocaine. The United States and Europe are responsible for 88 percent of global retail sales of cocaine, a market worth some US$70 billion in 2003(UNODC, 2005). The gross domestic product of the entire Caribbean was US$31.5 billion in 2004. (ECLAC, 2006). In other words, the value of the drug flows through the region may exceed the value of the entire licit economy. 7.2. Drug trafficking is associated with significant increases in crime, particularly violent crime and the use of firearms, as violence or the threat of violence regulates transactions in this market. In addition, drug trafficking is associated with money- laundering, trafficking of firearms, and corruption. Drug use is associated with increases in both violent crime and various types of property crime. For more on recent drug trafficking trends in the Caribbean, see Chapter 2; for regional policy responses to drug trafficking, see Chapter 10. 7.3. CARICOM's Regional Task Force on Crime and Security, speaks of a three- pronged strategy to international drug control: · Eradication or alternative development for producer countries (e.g. Colombia, Peru and Bolivia for cocaine) · Supply restriction through interdiction operations for transit countries (e.g. Caribbean states); and · Demand reduction for the main consumer countries (e.g. North America and Europe).(CARICOM, 2002) 7.4. This chapter examines interdiction efforts in the Netherlands Antilles. The Netherlands Antilles are an autonomous part of the Kingdom of the Netherlands. They are comprised of two groups of islands located about 900 km apart: 93 · Curacao and Bonaire, located in the south of the Caribbean, near the coast of Venezuela; · St. Eustatius, Saba, and St. Maarten, in the Eastern Caribbean.113 7.5. The Netherlands Antilles have a population of 183,000 people, (United Nations, 2004; United Nations, 2005)114 about 75 percent of whom live on the island of Curacao. The country is relatively well-developed, with a GDP per capita among the highest in the Caribbean, good infrastructure, and an economy based on tourism, financial services, and oil transshipment. 7.6. The Netherlands Antilles was chosen as a case study for two reasons. First, it is one of the Caribbean territories most afflicted by the drug trade and is among the world leaders in cocaine seizures per capita. Second, the government of the Netherlands Antilles, in collaboration with the Dutch government, has undertaken innovative and seemingly successful policies to interdict the supply of cocaine. 7.7. About 100 times more cocaine per capita was seized in the Netherlands Antilles than in the United States in 2004--over nine tons, or just under 50 grams of cocaine for every man, woman, and child on the islands.115 Some 60 percent of all the cocaine seized in the Caribbean in 2004 was seized in the Netherlands Antilles, and cocaine seizures increased dramatically between 2001 and 2004 (Figures 7.1 and 7.2).116 7.8. The increased flow of drugs though the region is believed to have a powerful impact on the local crime situation. The Netherlands Antilles authorities estimate that 75 percent of the crime on the islands is drug-related. On March 12, 2004, the Antillean government proposed a state of emergency due to the levels of crime afflicting the society, opening the door for both Dutch and Antillean military to participate in internal security operations. This was in response to a rapid increase in the number of crimes seen in Curacao. Drug-related murders increased from 12 in 2002 to 29 in 2003. Other islands experienced similar problems. The murder rate in St. Maarten went from 20 per 100,000 in 2001 to 47 per 100,000 in 2003.117 113St. Maarten is half of an island, the other half being St. Martin, a French territory. Aruba was part of the Antilles until 1986, when it became a separate country within the Kingdom of the Netherlands. While drug trafficking through Aruba used to be a major issue, it is much less so today. Aruba was removed from the US State Department's list of major drug producing and transit countries in 1999. The reason for this decline is unclear. 114World Population Prospects: The 2004 Revision and World Urbanization Prospects: The 2003 Revision, http://esa.un.org/unpp, accessed 25 July 2006. 115The Netherlands Antilles ranked eighth in the world in terms of gross cocaine seizures in 2004, just behind the Netherlands and just ahead of Bolivia and Brazil. 116 In addition to a possible increase in real volumes transiting the region, the dramatic rise in seizure figures is likely the result of a shift from an early focus largely on airport interdiction to intelligence-led operations against major traffickers, which only began to pick up around 2001. This was also the time that the Coast Guard of the Netherlands Antilles and Aruba (CGNAA) came on line, after being equipped with three cutters in 1999. 117Information on population and number of murders from the Central Bureau of Statistics Netherlands Antilles Statistical Yearbook as quoted on their on-line database: http://www.cbs.an/stat.asp. 94 7.9. As Figure 7.2 shows, the Netherlands Antilles have not always been at the epicenter of the global cocaine trade.118 Recently, a number of things have changed. Law enforcement efforts have made it less likely that cocaine shipments depart directly from Colombia than in the past, and more likely that they come from a secondary country, particularly Venezuela. Between much of the Venezuelan coastline and the rest of the world lie the islands of Curacao, Bonaire, and Aruba, so maritime cocaine shipments are likely to pass through the territorial waters of these countries. In addition, there is much commercial sea traffic between Venezuela and the Netherlands Antilles, which are just a few kilometers offshore.119 This traffic provides cover for drug shipments that can then be forwarded by sea or by air. Figure 7.1: Kilograms of Cocaine Seized in Caribbean Territories in 2004 Trinidad & Others Tobago 5% Bahamas 4% 5% 7 5 Jamaica 7 Netherlands 7 57 11% Antilles 1 1 60% 9 9 2 2 Dominican Rep. 15% Source: UNODC Delta database. 7.10. The islands' continued membership in the Kingdom of the Netherlands has also been important. While cocaine use in the United States is down since the 1990s, it has been increasing in Europe. The European portion of the Kingdom of the Netherlands is one of the two main points of entry for cocaine to Europe.120 The Netherlands Antilles have suffered from high rates of unemployment since the closure of the Shell refinery in 118In the early days of the Colombian cartels, large volumes of drugs were flown by private planes into the United States, their main destination. Improvements in radar response put an end to this, and shipments became increasingly maritime. Again, most sea routes from Colombia to the U.S. did not pass through the most populous areas of the Dutch Caribbean. 119Curacao is less than 65 km off the coast of Venezuela. Venezuela is the source of 52 percent of imports to Antilles. See Economist Intelligence Unit, "Netherlands Antilles Country Report." London: EIU, June 2006. 120The other being Spain. About 65 percent of the cocaine seized in Europe in 2004 was seized in Spain and the Netherlands. According to Europol, "Cocaine is also shipped from South America to Europe via the Caribbean Islands. This in part can be explained due to the historical links that exist between the Caribbean and some European countries, for example Curacao and the Netherlands ..." See Europol, "Drugs 2006." The Hague: Europol, 2006, p. 3. 95 the late 1980s, creating a pool of potential couriers with European passports (Central Bureau voor de Statistiek, 2006).121 Figure 7.2: Kilos of Cocaine Seized in the Netherlands Antilles, 1980-2004 10,000 9,025 9,000 8,000 7,728 7,000 6,000 5,000 4,000 3,000 2,532 2,455 2,000 2,124 1,302 1,000 965 906 1,043 624 710 639 590 23 364 - 11 41 57 54 25 132 140 111 18 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: UNODC Delta database. 7.11. A final factor in the remarkable seizure totals relates not to the existence of the flow of drugs but rather to their detection. Assisted by the Dutch Royal Navy and the Coast Guard for the Netherlands Antilles and Aruba (a common agency of the three countries in the Kingdom), Antillean law enforcement has intercepted some massive shipments in recent years, shipments that might have been missed by less well-resourced agencies, or only apprehended on arrival in Europe. Further, both the Antilleans and the mainland Dutch have also taken an innovative approach to stopping what had been a virtual stampede of couriers on commercial air flights. At least until recently, they have had a "drug focused" (as opposed to courier-focused) interdiction policy at the airports, a policy that is discussed further below. DRUG TRAFFICKING ON COMMERCIAL AIR FLIGHTS: AN INNOVATIVE POLICY RESPONSE 7.12. It is estimated that 30 tons of cocaine enter Europe on commercial air flights every year, and that Schiphol International Airport in Amsterdam has been one of the primary points of entry (Europol, 2006). In 2000, 4.3 tons of cocaine were seized at Schiphol (INCB, 2001), and by August 2001, the flow of cocaine from the Netherlands Antilles had reached crisis proportions. An innovative approach became necessary. The first priority became stopping the ingress of drugs, even if that meant identifying more couriers than could possibly be prosecuted. 121In 2005, unemployment was 15 percent on Curacao, including 44 percent among youth. 96 7.13. Toward this end, the "100% Control" policy was implemented, in which flights landing in Schiphol originating from the Dutch Caribbean, Suriname, or Venezuela are subject to extensive searches.122 Rather than attempting to scare off potential smugglers with the threat of incarceration, the Dutch approach was based on increasing the rate of interdiction to the point that smuggling becomes unprofitable. In other words, the focus was on the drugs, rather than the couriers, and was based on incapacitation, rather than traditional deterrence. Europol described the mechanics of the policy in this way: Crews, passengers, their luggage, the cargo and the planes are systematically searched. Couriers with amounts of less than 3 kg of cocaine are not detained, unless they are arrested for the second time or another criminal offense is involved. Instead, the drugs are confiscated and the smugglers are sent back. Couriers who have been identified are registered on a blacklist, which is provided to KLM, Dutch Caribbean Airlines and Suriname Airways (Europol, 2005, emphasis added). 7.14. Detecting the couriers is not easy, because most carry the cocaine in their intestines. These "swallowers" (slikkers) ingest little balls (bolletjes) of cocaine, an average of 90 per courier, usually wrapped in two latex glove fingers and coated in wax (National Ombudsman, 2006).123 If one of these bursts, the courier usually dies of an overdose. Consequently, this form of couriering appeals most to people who are reckless, desperate, or ignorant. In other words, it appeals to people who are difficult to deter with the threat of incarceration in a European prison. 7.15. The first recorded case of cocaine swallowing in the Netherlands Antilles was in 1983. This form of couriering appears to have been exploited early on by West African organized crime groups, which have utilized body couriers to traffic drugs between markets all over the world. The great advantage of using commercial flights is that any criminal entrepreneur with a small amount of capital can try, and this makes the method attractive to the loose, flat networks that typify West African organized crime. The West Africans soon abandoned personally couriering the drugs in favor of hiring people less likely to be searched, starting with Brazilians. Today, East Europeans feature prominently in the courier ranks. However, arrest statistics of the Dutch Royal Military Police, responsible for enforcement at Schiphol Airport, show that 46 percent of those arrested for drug offenses in 2004 were Antilleans (Trimbos Institute, 2005). 7.16. West African networks also pioneered the use of the "shotgun" approach, where multiple couriers are placed on the same flight with the hopes that a few will get through. Traffickers recognize that law enforcement authorities can only arrest a limited number of couriers on any given day, as each requires extensive paperwork and subsequent court time. Once a few are detected, attention turns away from the rest of the passengers. By sending several couriers on one flight, traffickers are virtually guaranteed that some of 122In practice, flights from Ecuador and Peru are also subject to 100% control, as all flights from these countries stop over in the Dutch Caribbean. 123It has also been alleged that cat or goat intestines have been used as wraps as they are apparently more difficult to detect. 97 their drugs will get through. At least, that was true before the 100% Control strategy was put in place. 7.17. Drug smuggling remains an attractive option for young islanders. While Antilleans are Dutch citizens and thus benefit from extensive social protections, just getting by may not be enough for some young people. Caribbean youth often speak of the "easy money" to be gained through the drug trade, as compared to the near impossible employment situations many face (World Bank, 2003a). Participation in the drug trade offers a tempting way to earn some relatively significant income quickly. 7.18. Under 100% Control, suspected couriers are taken aside and questioned. Once officials have a firm basis for suspicion, the suspects are given the option of proving their innocence by submitting to a body scan. Alternatively, they are kept in custody until their intestines empty themselves, whereupon culpability is positively ascertained. The Dutch authorities even offer a receipt for the drugs taken, so that couriers can explain the loss to their employers. 7.19. The strategy appears to have been remarkably successful. Between 1 January 2004 and 1 April 2006, just under 4,000 flights were examined, 6,147 couriers were identified and 7.5 tons of cocaine were seized,124 accounting for more than half of the cocaine seized at Schiphol airport. This amounts to 1.2 kg per courier, over 70 percent of whom carried the contraband internally (National Ombudsman, 2006). Figure 7.3 shows the decline in the number of couriers detected arriving from Curacao at Schiphol. While the controls have remained quite consistent throughout, the number of couriers detected has dropped dramatically, from a peak of 463 couriers in the second quarter of 2004 to less than 20 per quarter today, a 96 percent reduction. 7.20. The actions at Schiphol are only part of the story.125 The approach was actually initiated at Hato International Airport in Curacao in 2002, and today is being extended to other airports in the Netherlands Antilles, particularly St. Maarten. Here, all passengers are screened and those identified as potential couriers are questioned. Those identified as couriers are given the option of proving their innocence, and thus being allowed to board the plane, by submitting to an x-ray scan. Only about one in ten suspected couriers elect to be scanned, and some of these are proven to be guilty. Those who refuse are not allowed to board the plane and are not given a refund on their ticket, so failure to board is tantamount to an admission of guilt. 124To put this quantity in perspective, it is more cocaine than was seized annually in any but the top 10 cocaine-seizing countries in 2004. 125The importance of the actions on the Antilles side is demonstrated by the lesser success of the 100% Control strategy for couriers originating from Suriname, an independent country where cooperation has not been as close with the European Dutch. 98 Figure 7.3: Couriers Detected Arriving at Schiphol from Curacao, by Quarter 500 463 450 415 400 399 350 343 300 250 200 161 150 130 100 40 50 21 18 17 0 1/2004 2/2004 3/2004 4/2004 1/2005 2/2005 3/2005 4/2005 1/2006 2/2006 Source: Netherlands Ministry of Justice. 7.21. When this procedure was initiated, up to 100 people a weekend were denied boarding. Between mid 2002 and 2004, at least 13,000 people sacrificed their air ticket rather than face the body scan (Bureau for International Narcotics and Law Enforcement Affairs, 2005). 7.22. Thus, the remarkable seizures at Schiphol were just the residual that evaded the initial screening. On this basis, the authorities estimate that between 80 and 100 couriers per day were passing through the airport in 2003. This was cut to 10 a month by October 2005 (Bureau for International Narcotics and Law Enforcement Affairs, 2006). This reduction was accomplished not by draconian sentencing of the mules, but by making the route unprofitable to the traffickers organizing them. 7.23. The impact of this screening on organized crime was manifest in the violent response the airport team frequently encountered in the early days of the program. In 2003, they suffered at least seven firearm attacks in the airport, as well as threats and assassination attempts while off duty. At one stage, the body scan team was forced to suspend operations due to the threats they had received. 7.24. On those occasions where couriers are prosecuted and convicted in the Netherlands Antilles, they are given, in some cases, an alternative to incarceration: passport forfeiture. As part of a sentencing agreement, convicted couriers surrender their passport for a specified period of time (up to three years). The Netherlands Antilles government has collected over 800 passports in this way. As with the 100% Control strategy, the intent is to incapacitate the couriers without the costs of incarceration. 99 7.25. While displacement effects would need to be tallied to properly evaluate the impact of this intervention, the 100% Control strategy has apparently disrupted what was once a major trafficking route. Of course, the resources required to individually scrutinize every flight and passenger come at a cost, but these expenses are likely to be less than those of processing and incarcerating an endless stream of hapless mules. 7.26. The 100% Control strategy is not without its critics. The Dutch National Ombudsman126 recently issued a report critical of the program in several respects (National Ombudsman, 2006). The Ombudsman regards as particularly objectionable the detention and search of innocent people, which comprise about one quarter of all suspects, and their treatment even when exonerated.127 Earlier practices, which involved body cavity searches, were abandoned on the advice of the Ombudsman. The policy has also been attacked by national leaders from the Dutch Caribbean. Suriname's President Ronald Venetiaan called the policy "inhumane, barbaric, and primitive."128 It was also loudly condemned by Anthony Godett, who was set to become Prime Minister of the Netherlands Antilles before he was incarcerated on fraud and corruption charges. 7.27. This places the Dutch in a difficult position. They have a strategy that apparently works, and that is intended to be both more humane and more cost effective than incarcerating those that risk death by swallowing kilograms of cocaine. To be effective, however, a large number of people need to be searched, including some who may prove to be innocent. A more restrictive approach would risk failing to capture a sufficient share of the drug flow to make the route unviable. 7.28. Abandoning the approach altogether is unthinkable, as there is evidence that the former torrent would resume immediately if there were a break in policy. For example, on December 26, 2005, two flights from Curacao to Schiphol were offered to accommodate the flow of returning tourists, one of which was scheduled uncharacteristically early in the morning. Apparently, couriers assumed that controls would be lighter or absent for the earlier flight, but had an unpleasant surprise when they arrived at Hato International. Dozens of no-shows left the fully-booked flight half empty.129 7.29. Since Schiphol airport is now seeing only a courier a week from Curacao, the detention of innocent people is likely to be minimal, but re-routing may mean that other airports are likely to adopt a similar strategy,130 so the challenge of creating an even- handed and less controversial approach remains. There is a need for further research on the costs and benefits of the 100% Control system, and frank discussion of the ways it 126This office is tasked with investigating complaints against civil servants on behalf of the public. 127As of mid-2006, some 2,271 people proved to be `negatieve slikkers' out of 8,576 people searched, according to information provided by the Netherlands Ministry of Justice. 128Cairo, I, `Suriname to complain against Dutch drug inspections'. Caribbean Net News, 20 March 2006. 129Anecdote relayed in an interview with the Dutch Ministry of Justice, 27 July 2006. 130The Netherlands is not the only country in the world to receive flights from the Antilles, or to find couriers on those flights. For example, Canada arrested at least eight traffickers carrying over 100 grams of cocaine apiece on commercial air flights from the Antilles in 2004, according to the UNODC Major Seizures Database. 100 could be improved. The fact that some difficulties have been encountered with implementation does not mean the central principle--drug, rather than courier, interdiction--has been invalidated. 7.30. Of course, commercial air flights are only one of the vectors through which cocaine transits the Netherlands Antilles, and not necessarily the most significant one. The 100% Control approach has seized 7.5 tons of cocaine in two years, but multi-ton seizures can be made in a single instance of maritime interdiction. Since the share of sea shipments seized is almost certainly much less than the share of air couriers apprehended, it is likely that the vast majority of the cocaine traffic is maritime. MARITIME TRAFFICKING 7.31. The first boat trip cocaine encounters on the Netherlands Antilles route is the short passage from the coast of Venezuela to Curacao, about 65 km away. This trip is essential whether the drug will ultimately be trafficked by sea or by air. Traditionally, go- fast boats carrying between half a ton and a ton of cocaine have been used. But in 2004, the authorities noted a shift to smaller shipments (50-500 kg) in smaller, slower boats, including fruit and fishing boats (UNODC, 2004). 7.32. On arrival in the Netherlands Antilles, cocaine destined for onward maritime shipment is generally consolidated. An estimated 230 tons of cocaine enter the European Union annually via maritime shipments and air freight, an unknown share of which transits the Netherlands Antilles (Europol, 2006). Large seagoing vessels have been used to bring this contraband to Spain, increasingly via West Africa. The extent of containerized traffic of narcotics to Europe is debated. There are few concrete cases, but U.S. authorities suggest that this is an active trafficking vector (U.S. Department of State, 2006). 7.33. The United States also receives cocaine from the Netherlands Antilles, often trafficked via go-fast boats and in the holds of leisure sea craft from St. Maarten to Puerto Rico, but this flow is not considered a significant source of the United States' supply. However, law enforcement cooperation between the United States, Colombia, and the Netherlands Antilles has led to some significant drug arrests in recent years. 7.34. Detecting maritime trafficking requires intelligence work and a dedicated interdiction force. In 1996, the Coast Guard for the Netherlands Antilles and Aruba (CGNAA) emerged as a distinct entity after acquiring three cutters designed for combating drug traffickers. It continues to work with the Royal Navy of the Netherlands (RNN) in making major seizures. In 2004, the Netherlands Antilles saw their largest seizure ever when some 2.5 tons of cocaine and 28 kg of heroin were seized. But since that time, numerous large maritime seizures have been made: · On June 25, 2005, 1.8 tons of cocaine was seized by the CGNAA and RNN (Netherlands Ministry of Justice, 2006). · On July 21, 2005, a Venezuelan craft carrying 792 kg was seized (Netherlands Ministry of Justice, 2006). 101 · On August 27, 2005, a joint U.S./Dutch effort seized over two tons of cocaine from the Bolivian flagged M/V Sea Atlantic (U.S. Drug Enforcement Administration, 2005). 7.35. While final seizure figures for 2005 have not yet been released, these three seizures alone would amount to about half the cocaine seized in 2004, a record year. 7.36. Of course, the Netherlands Antilles and Aruba are situated between Venezuela and Europe, just off the Colombian border. This location means that shipments directly from Venezuela to Europe often pass through Antillean/Aruban waters incidentally. In short, the CGNAA and the RNN are doing a lot of Europe's interdiction work close to the source. These seizures may have little to do with the Netherlands Antilles or the Antilleans. 7.37. All indications are that Venezuela will continue to grow as a conduit for cocaine, and this is likely to mean that maritime seizures in the Netherlands Antilles will remain high. As the Colombian government continues to make progress, those elements of the FARC and the AUC committed to conflict or criminal incomes are likely to withdraw into Venezuela. The small Caribbean countries will need international support if they are to continue to form the front line in maritime cocaine interdiction. POLICY IMPLICATIONS 7.38. Looking strictly at the seizure figures, one might see the disproportionate use of the small Netherlands Antilles for drug trafficking as an indictment of Dutch drug policy. The reality is more complicated than that. The Netherlands Antilles are somewhat unique in their geographic vulnerability, particularly with regard to maritime trafficking. In essence, they constitute Europe's (and, to a lesser extent, the United States') first line of defense against drug shipments coming from the Venezuelan coast. The high seizure figures are largely a product of remarkable police work, and should be commended, not condemned. 7.39. The fact that the Netherlands Antilles are preferred by air couriers is another matter. This is undoubtedly tied to the fact that the Netherlands is a key drug distribution center for Europe, which in turn is tied to the perception that the Dutch are soft on drugs. But here again the Netherlands have risen to the challenge and, at considerable expense and effort, have managed to stem what was once a major source of cocaine supply to Europe. Not only that, they managed to do so in a way that did not destroy the lives of the troubled youth and addicts that comprise the corps of international drug mules. 7.40. Of course, even eliminating the Netherlands Antilles as a drug transit area altogether will not stop the flow of drugs, and displacement effects are inevitable. French authorities are already asserting that Charles de Gaulle airport is increasingly being targeted by air couriers, and lesser airports are also seeing increased drug traffic. In August of 2006, almost 40 kg of cocaine was seized from French and Dutch nationals in six separate incidents at Orly airport in France. Cocaine seizures in nearby Trinidad and 102 Tobago have skyrocketed since the program really began to take off in 2003, although whether this is strictly a displacement effect is unclear. Figure 7.4: Cocaine Seizures in Trinidad and Tobago 2,000 1,892 1,800 1,600 dezies 1,400 eniacocfo 1,200 1,000 800 821 kg 600 590 400 391 200 179 203 137 173 173 94 95 57 71 78 - 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Source: UNODC Delta database. 7.41. Despite their success, the Dutch seem intent on disavowing the drug-centered interdiction approach, having faced considerable international criticism before all the numbers were in. Today, they are quick to insist that all couriers will be prosecuted, but often neglect to mention that this is only possible because the numbers have declined so radically, due in no small part to the fact that not every courier was prosecuted in the past. This reluctance to promote drug-centered interdiction is unfortunate, because the 100% Control experiment represents an innovation worthy of further promotion and testing. 7.42. Of course, Caribbean countries may object to the intense scrutiny of passengers arriving from their countries. But the alternative, involving the arrest and prosecution of many Caribbean nationals for drug trafficking, may be even less palatable. Officials positively identified over 6,000 couriers arriving at Schiphol from the Dutch Caribbean in less than three years, at least half of whom were Antilleans. In the absence of a drug- focused policy, all these people might be in Dutch jails today. 7.43. Focusing on the drugs rather than the couriers is a powerful approach because it defeats the "shotgun" technique, where traffickers are willing to sacrifice an endless stream of gullible couriers so long as sufficient quantities of drugs arrive for them to make a profit. It has proven itself to work in the case of the Netherlands Antilles, and could be tested in other contexts, including other Caribbean countries suffering from drug transshipment. Once there is clear evidence that a particular routing is being used by drug mules, measures similar to the 100% Control approach could be taken at both source and destination airports. Drug-focused interdiction bypasses the couriers, who, willing to risk their lives for a few thousand Euros, are unlikely to be deterred by threat of incarceration. 103 Instead, it speaks directly to the traffickers in the language they understand: it renders drug routes unprofitable. 7.44. The Antillean example also highlights the need for cooperation between Caribbean transshipment countries and destination countries in maritime interdiction, as called for by CARICOM's Regional Task Force on Crime and Security (CARICOM, 2002.) Given that small Caribbean countries have inherent limitations in law enforcement capacity, a little assistance can go a long way. 7.45. Of course, this case study does not deal with all elements of the drug puzzle in the Caribbean; the Caribbean Regional Task Force on Crime and Security identified several challenges beyond successful interdiction: · On the demand side, to reduce demand and treat and rehabilitate drug abusers. · On the supply side, to find ways to counter the attraction of cultivation of illicit crops as a means of livelihood. · In terms of rehabilitating those convicted of drug possession, to introduce a penal system that places greater emphasis on uniform, non-custodial sentences for drug abusers and minor drug offenses, with the aims of promoting rehabilitation, and reducing prison overcrowding. 104 8. THE CONTRIBUTIONS OF CRIMINAL JUSTICE SYSTEMS TO THE CONTROL OF CRIME AND VIOLENCE: A CASE STUDY OF JAMAICA AND THE DOMINICAN REPUBLIC 131 Criminal justice reform is a key component of broader multi-sector approaches to reducing crime and violence in the Caribbean. In recent years, traditionally organized, criminal justice institutions have been faced with demands that they respond more effectively to increasing levels and new forms of crime and violence. Ongoing debates over the ability of criminal justice organizations to reduce crime levels make it clear that their impact can be overstated and that improving sector performance at best affects the proximate--and not the fundamental--causes of crime. Nonetheless, in response to citizen demands many nations in the Caribbean region are attempting to reform and modernize sector institutions, combining higher budgets (to allow increased staffing and the creation of more service units) with improved management, upgrading of personnel and personnel administration, new technologies, and the introduction of new operating procedures. Most of this work has been done with police, courts, prosecutors, and public defenders; very little has been attempted with prisons despite their potential role in deterring crime and rehabilitating offenders. There is also an increasing focus on developing management information systems and performance indicators for better problem diagnosis, tracking of system outputs, monitoring reform programs, and increased accountability to citizens. This chapter reviews these developments in the context of two of the largest countries in the region: Jamaica and the Dominican Republic. It also reviews some of the statistics generated as a means of tracking recent trends and gaps in reform strategies. Although most reform efforts have focused on individual institutions, two important lessons emerge: (i) the need to pursue better coordination among institutions, including the introduction of information systems capable of tracking systemic performance, and (ii) the desirability of linking criminal justice reforms to a broader, multi-sector strategy of crime and violence prevention. INTRODUCTION: CRIMINAL JUSTICE REFORM IN CIVIL AND COMMON LAW SYSTEM 8.1 Although there has been a tendency in the Caribbean to over-rely on law-and- order responses to crime and violence (as opposed to other, more prevention-oriented approaches--see Chapter 9), the criminal justice system remains an essential tool in crime and violence control and, to some extent, in their prevention. Academics and researchers specializing in the topic remain divided as to the real impact of the criminal justice system on reducing crime. As James Q. Wilson--an early proponent of the strict responses to crime and the author of the original "broken windows" thesis--has written, 131This chapter is based on a background paper by Todd Foglesong (Senior Program Associate at the Vera Institute of Justice in New York and Senior Research Fellow at the Kennedy School of Government) and Christopher Stone (Daniel and Florence Guggenheim Professor of the Practice of Criminal Justice, Kennedy School of Government, Harvard University). The authors are grateful to Andres Rengifo and Glendene Lemard for research assistance and reviews of drafts of the background paper. Additional material on Jamaica was provided by Florence Darby (independent consultant), and on the Dominican Republic by Linn Hammergren (World Bank). Linn Hammergren and Stephanie Anne Kuttner drafted this chapter. 105 "There is no silver bullet that will reduce crime, much less eliminate it...[N]o one can satisfactorily explain changes in crime rates "(Wilson, 2004). 8.2 This uncertainty, however, cannot be an excuse for inaction, especially for governments facing increasing levels of crime and violence and citizen demands that they do something about them. This is still more critical in countries whose criminal justice systems have languished unattended for decades, lack the technical and human resources to respond to less traditional types of crime, have not adopted new standards of transparency and accountability, or have themselves been penetrated by political and occasionally, criminal elements. 8.3 Criminal justice reform is part of a broader law and justice reform movement that began twenty years ago in the Latin American region, and somewhat more recently, in Eastern Europe (Hammergren, 2002). Lessons can be drawn from this broad international experience, as well as from experiences in the Caribbean (IADB and Caribbean Group for Cooperation in Economic Development, 2000). 8.4 One frequent observation is that criminal justice reform can be very difficult, if not impossible, absent sufficient political will. Stakeholder support that reinforces the political will to reform is equally indispensable and thus a good understanding of reform supporters and potential detractors, both within the criminal justice system and society more broadly, is key to designing the appropriate reform strategy. A second important lesson is that success of these efforts depends on aligning the work of the different institutions within the system, and of the system as a whole with a larger, multi-sector crime and violence prevention strategy. Until recently, these two elements tended to be overlooked, occasionally with very unfortunate results.132 8.5 In the Caribbean region, the two major legal traditions or families, common and civil law, have taken somewhat different approaches to criminal justice reform. In the civil law systems (and especially those of Spanish and French origin), the initial emphasis was on modernizing criminal procedural codes to effect a shift from inquisitorial to more accusatory systems and incorporate more rights for the defendant. The due process emphasis was a logical reaction to the abuses of prior authoritarian regimes and also 132A donor-sponsored effort to recreate a Haitian police force in the 1990s, for example, floundered in part because of inadequate attention to other elements of the criminal justice chain. Courts and prosecution continued to operate much as before in processing cases ­ slowly, inefficiently, and impaired by politicization and corruption ­ and moreover provided no more check than they ever had on police abuses. Prisons, which received still less attention, were quickly overcrowded by those arrested by the police, meaning that inmates suffered from inhumane conditions, went long periods without being seen by a judge, and occasionally protested by staging riots and further damaging already inadequate facilities. This truncated reform has been reviewed critically by many experts, including some who were involved in its implementation. For a summary of the initial process, see Rachel Neild, "La Reforma Policial en Haiti: Un Triunfo sobre la Historia," in Lilian Bobea, Soldados y Ciudadanos en el Caribe, Santo Domingo: Flacso, 2002, pp. 285-308. Although, as the title suggests, Neild does not discount the successes of the effort, she is quite clear that failures to attack related problems like political intervention and an unreformed judiciary were already (in 1998, when the article was written) threatening to undercut the accomplishments. 106 followed trends already underway in continental Europe.133 The incorporation of crime control objectives came slightly later, so that some of the initial legal changes occasionally appear to conflict with the pursuit of crime control. While there is no inherent contradiction between due-process and crime-control in procedural codes, codes written to advance only one of these objectives can undermine progress in the other.134 More importantly, it is increasingly recognized that law reform does not automatically advance either end, and thus that more attention must be paid to rationalizing organizational structures, improving internal management and oversight, human resource development, and the introduction of new technologies and technical skills. Progress here has been spotty, partly because of the financial investments required, and partly because of the resistance of interests vested in the inefficiencies of the traditional systems. 8.6 In the common law countries, legal change has been less emphasized and more importance has been given to upgrading and strengthening institutional capacities. As in the civil law countries, there is an emerging consensus on the importance of integrated reform programs that simultaneously improve and coordinate the functioning of different actors and institutions within the criminal justice system--that is, to ensure linkages between police reform, reorganization of prosecution agencies, strengthening of the judiciary's capacity and independence, and modernizing correctional systems. However, as in the civil law nations, effecting coordination in fact is far more difficult than espousing it in principle. 8.7 Traditional criminal justice systems from both legal families react to crime by incapacitating offenders though incarceration, sending a deterrent message to potential offenders, and providing the society with a sense of justice. The approach is usually popular with the public, but as even sympathetic experts like Wilson note, its efficacy is subject to dispute; it still requires sophisticated strategies, inter-institutional coordination, and strict oversight; and absent these conditions, it can lend itself to corrupt and abusive application. 8.8 More modern systems stress additional elements: (i) preventive policing; (ii) more strategic approaches to prioritizing crimes for investigation and prosecution; (iii) rehabilitation of prison inmates and re-entry programs to reduce recidivism; and (iv) use of alternative sentencing, fast track procedures, and conciliation with the victim for minor offenses. All of these additions are intended to make more efficient use of resources and 133Throughout Latin American and the Caribbean, entrenched authoritarian regimes often used the police and other criminal justice agencies as a means of repressive control. Even after a democratic transition, the structures and practices introduced for this purpose may be hard to eradicate, but unless they are reversed, other reform activities are unlikely to work and may even have perverse consequences. Community policing and other preventive tactics in the hands of a corrupt, irresponsible police force may be an invitation to harassment of citizens. More independent judges insufficiently monitored by their own institutions may remain susceptible to bribes and favoritism. 134As just one example, many of the early reformed codes stipulated that within ten days of the initiation of an investigation, the person(s) targeted had to be informed so that they could prepare their defense. While for simple crimes this may pose little problem, for those of a more complex nature (e.g. drug trafficking, money laundering, and grand corruption), the provision could stop the investigation in its tracks. Poorly drafted provisions about police-prosecutorial coordination also commonly aggravated conflicts between the two institutions. Hammergren, op cit. 107 to maximize the impact on discouraging crime before it happens or on limiting its further effects. 8.9 The following sections first describe the criminal justice reforms currently under way in the Dominican Republic and Jamaica. This brief overview provides the context for a subsequent comparative performance analysis based on statistics recorded by the two countries. A final section discusses governance structures that are conducive to successful reform. COUNTRY CASES: TWO APPROACHES TO CRIMINAL JUSTICE REFORM The Dominican Republic 8.10 The Dominican Republic, while one of the few civil law countries in the Caribbean, is a quintessential example of the broader Latin American justice reforms. Trujillo's thirty-one year dictatorship (1930-1961) was accompanied by a centralized, ad hoc domination of all governance institutions, including those in the justice sector. With the advent of democracy, this system was quickly replaced by party-based, patronage politics in which governmental elites used public employment as a means of consolidating coalitions and rewarding followers. 135 Low official salaries and massive turnovers at the end of every administration increased the incentives to seek irregular supplements and to dedicate little effort or time to one's official job. 8.11 In the justice sector, the results were predictable--inefficiency, high levels of corruption and abusive practices, poor service delivery, and low public confidence (Hasbun et al., 1999). In the 1990s, public discontent with the situation, spearheaded by a local NGO, FINJUS, provoked a series of reform measures (Salcedo, 2003).136 The first set involved: the introduction of a new system for appointing Supreme Court justices,137 the Court-supervised renewal of a good part (70 percent) of the rest of the bench via transparent, merit-based examinations of seated members and aspirants; the creation of a judicial career path; and, increased funding to improve judicial salaries and finance equipment, infrastructure, and training programs. 8.12 A second, more far-reaching reform was the adoption of a new criminal procedure code in 2002. In line with regional trends, the code featured more accusatorial 135Low-level police officers were one exception. Since their jobs are dangerous and poorly paid, there is not an abundance of would-be recruits. Higher-level officials, however, were frequently chosen for their political connections, not their probity or inclination to encourage better police work. For a discussion of the general problem, with some reference to the police and judiciary, see World Bank, Dominican Republic Public Expenditure Review: Reforming Institutions for a More Efficient Public Expenditure Management. Washington, D.C.: the World Bank, Report No. 23852-DO, March 15, 2004. 136For a discussion of the reforms and of the complaints leading up to them. 137Justices had formerly been appointed by the legislature. The change was introduced by constitutional amendment in 1994 but not put into effect until 1997. It established a judicial council, headed by the national president and with members from other branches of government and the private bar, which met solely for the purpose of filling vacancies in the court. The council's deliberations and the list of candidates were made more transparent, thereby, at least in the first round, discouraging appointments based solely on political allegiance. 108 elements and a greater role for public prosecutors and defense. With donor assistance, the old public defense system, based on court-appointed lawyers, was converted to a formal career with measures to supervise public defenders' performance. This eliminated a major source of corruption; many former defenders were known for taking and offering bribes and charging for theoretically "free" representation of indigent clients. Donors have also been active in supporting programs to improve and professionalize public prosecution, but here progress has been far slower. A career system has only been partially implemented for ordinary prosecutors, with the Attorney General and the district prosecutors (who oversee other members of the institution in their respective geographic jurisdictions) still appointed and removed by the Executive. 8.13 With strong support from local NGOs, a new police law was enacted in 2004. Its aim was to demilitarize the national force, promote professional development, improve internal oversight, and increase accountability to civil society (Bobea, 2003; Grullon, 2003). Unfortunately, the law's implementation has faced resistance from those within and outside the police force interested in retaining the traditional practices. While there have been some interesting experiments with new policing techniques (see Box 8.1), the institution's public image remains poor, and there are strong indications of continuing police involvement in criminal activities and in shaping investigations to favor the politically and economically powerful.138 8.14 The Dominican criminal (and wider) justice reform thus remains half-completed. There are progressive elements in all the basic institutions who are interested in promoting further change, but more efforts are needed to finish the job of rooting out the corrupt and inefficient (Participación Ciudadanía, 2004).139 With the exception of the largely untouched prison area and public defense, budgets are arguably high enough and the sheer quantity of employees sufficient to do better, but there are problems related to how resources are used, the quality of professional and support staff, and the incentives shaping their behavior. 138For a discussion of police corruption, see DR1, January 20, 2005 citing comments by a delegation of experts from the John Jay College of Criminal Justice, New York and from FLACSO, http://www.dr1.com/news/dnews012005.shtm. The former notes that the "difficulties include the low wages paid to police officers, the bad reputation the force has for rampant corruption in the ranks, and lack of facilities to do the job." FLACSO's accompanying comment is that "crime in the DR is closely linked to the corruption levels within the police and the unskilled police officers" See also the U.S. State Department country report on human rights for 2005 (http://222.state.gove/g/drl/rls) which notes problems originating in a failure to vet police recruits and World Bank Country Memorandum on the Dominican Republic which notes (p, 156) that 48 percent of firms surveyed reported paying bribes to police. http://www- wds.worldbank.org/external/default/main?menuPK=64187510&pagePK=64193027&piPK=64187937&the SitePK=523679&menuPK=64154159&searchMenuPK=64258544&theSitePK=523679&entityID=000090 341_20061122101801&searchMenuPK=64258544&theSitePK=523679 139On court and prosecutorial inefficiency in handling corruption cases, FINJUS has been particularly outspoken, noting that as of early 2005, and despite numerous high level scandals (most notably Baninter and PEME, the job-creation program), there had yet to be a single conviction. See www.redinter.org/Democracria-al-dia/41135. 109 8.15 Both the criminal procedures and police reforms arguably placed too much faith in the ability of law to alter behaviors and to overcome traditional practices. Slowness in reforming the public prosecution and the police force has so far impeded more radical change. There are also unfortunate signs of some backsliding in the judiciary, especially in the handling of corruption cases involving political and economic elites (Participación Ciudadanía, 2004). Cases have been dismissed, delayed or had charges reduced because of what the judges term prosecutorial incompetence and what the prosecutors claim is judicial collusion. As this and the troubled relations between police investigators and prosecutors suggest, the reform has also paid insufficient attention to the need to coordinate activities across institutions. 8.16 Thanks in part to donor support, the police, judiciary, prosecution and defense have established systems to track activities in each institution. The court's actions in this area in fact date to before the reforms. However, it is not evident that the resulting information is being used to monitor reform progress and detect problems, and there has so far been no effort to link the databases of the separate institutions so as to be able to assess their joint handling of cases in which they all participate. Although the Dominican Republic has recently established a national council to develop a multi-sector strategy for dealing with crime, the country's track record with such special committees has not been positive. Often composed of agency heads or prominent local experts, they frequently possess moral authority, but few resources and technical staff to support their work. Jamaica 8.17 As a former British colony and current member of the Commonwealth, Jamaica already has a common law, accusatorial justice system. Pending the establishment of a Caribbean Court of Justice, its court of last resort is the United Kingdom Privy Council. Its Appeals Court is the highest local body. At the next level, its Supreme Court includes a panel to try major felonies, and another to handle gun issues. The more numerous magistrate courts have criminal divisions to handle minor crimes. Their approximately fifty members also serve on a series of specialized courts for traffic, family and juvenile matters. Jamaica also has roughly 8,000 justices of the peace. As they are not lawyers, their judicial role (for example, sitting on the 14 Petty Sessions Courts) is restricted to handling very minor disputes. Their more common functions are clerical or notarial, signing passport applications, certifying the identity of applicants for drivers licenses and the like. 8.18 Contrary to practices elsewhere in the Commonwealth Caribbean, the 1962 Constitution created a separate prosecution body, the Director of Public Prosecutions (DPP), thus eliminating the use of police prosecutors for criminal cases in any of its courts. The 7,200-member Jamaica Constabulary Force (JCF) is responsible for most policing. Its efforts can be supplemented in some areas by the Jamaica Defense Force (JDF). The country has eight adult prisons and four juvenile detention facilities, all of which are overcrowded and in physical disrepair. Segregation of other groups (e.g. the mentally ill) needing special attention, rehabilitation and reintegration programs, and programs for young offenders are either deficient or non-existent because of resource constraints and popular opposition to putting more funds into prison reform. 110 8.19 Jamaica has faced a high and increasing level of violent crime for the past thirty years, with drug trafficking and organized gangs playing a major role.140 While crime has traditionally been highest in urban areas and especially in the inner-city slums of the Kingston metropolitan area, police contend that their anti-crime measures have forced criminals into rural areas where rates are rising as well. Prison crowding and public outrage are also increased by the delays in investigating and prosecuting criminal cases. The small court and prosecution systems are swamped, and the non-use of plea bargaining and alternative dispute resolution (ADR) for minor crimes tends to delay trials inordinately. While corruption, especially in the police, is sometimes blamed for the perceived failure to bring the guilty to justice, more important factors appear to be the inadequate numbers of judges and prosecutors, poor police and general system image, especially in low-income communities, and a consequent tendency for citizens to take more direct methods in hand (Darby, 2006).141 Poor community relations have been aggravated by what the government calls a "suppression of crime culture" originating in the 1974 Suppression of Crimes Act. The act gave the police extensive search and seizure powers without a warrant. Only repealed in 1993, it is said to have influenced an entire generation of police officers who felt empowered to act without due process constraints. 8.20 Over at least the past decade, the government has introduced a number of measures to improve overall efficacy and efficiency while eliminating abusive practices. Donors have been instrumental here, although also facing complaints of a failure to coordinate their own actions. Citizens and NGOs have also been invited to participate in these efforts. As regards expediting court cases, measures include the establishment of night courts, a drug court, and regional gun courts, as well as the expansion of the number of magistrates; the enforcement of the rule that Clerks of Courts (prosecutors in the Magistrates Courts) be attorneys, the inclusion of a trained court administrator in each court, training of stenographers for Magistrates Courts, a 2000 amendment to the Criminal Justice (Reform) Act of 1978 that allows judges to impose non-custodial sentencing options, a Criminal Justice (Plea Negotiation and Agreements) Act (2004) and, in civil justice, the development of new Procedures Rules, case management systems, and ADR mechanisms. The government proposes to extend the case management system to criminal cases as well. In January 2006, the cabinet announced a five-year program to reform the justice system, to coordinate its own actions as well as donor assistance. Although not legally mandated, both the courts and the DPP have begun to provide Parliament with periodic reports on their activities. 8.21 On the police side, the revocation of the Suppression of Crimes Act represents a first step. In 1999, the Police Public Complaints Authority (PPCA) was established. While functioning, it investigates only a fraction of fatal shootings by police, in a country with one of the highest rates of police killings in the Western Hemisphere. Amnesty International reports the PPCA head as saying that "he did not consider it appropriate or possible to hold the Jamaican police accountable to the same standards as other countries, 140A poll done by Stone Polls in August, 2006, found 64 percent mentioning crime as the most important problem, followed by joblessness (20.7 percent) and leadership (1.7 percent) 141As Darby notes, there are problems with police corruption as well, especially as regards collusion with drug gangs. 111 due to the high level of gun-related crime in the country," (Amnesty International, 2001). The government has also introduced a number of initiatives to partner with law agencies in other countries as a means of combating organized criminal gangs, improving security in marginalized communities by providing training to and facilitating partnerships with community members, strengthening border control of illegal drugs, creating community action committees to encourage community-based crime control, and creating a Police Civilian Oversight Committee. Like the reforms in the other sector agencies, most of these are too new to evaluate their chances of success or even sustained implementation. 8.22 Although sheer quantities of human and financial resources are often less of a problem than claimed, it would be hard to contend, given the levels of crime and violence, that they are not problematic in Jamaica. Existing resources might be more efficiently deployed and controlled, and there is certainly room for managerial and legal change to combat court delays. However, the low judge and prosecutor to population ratios (3.2 and 3.1 per 100,000 inhabitants as compared to the Dominican Republic's 7.0 and 8.0 respectively and a Latin American average for judges of 8.1), while characteristic of the English style common law system, seem inadequate to the existing challenges either in processing cases or detecting and prosecuting police malfeasance. 8.23 There is also an interesting lesson here for the Dominican Republic in the failure of Jamaica's accusatory system to meet citizen performance standards, especially since both countries currently confront complaints of uncontrolled police abuse, poor community relations, and low clearance rates for major crimes. Clearly an accusatory system does not resolve these problems automatically, and more direct actions must be taken. A QUANTITATIVE EVALUATION OF SYSTEM PERFORMANCE: PERFORMANCE STATISTICS AND INDICATORS 8.24 The somewhat disappointing returns to many standard reform initiatives have prompted researchers, policy makers, and practitioners of legal and judicial reform to search for better ways of assessing the performance of justice systems and tracking progress in reform programs (World Bank, 2005b). Their universal conclusion is the need to improve the statistics collected by organizations on their own operations and to encourage the development of management information systems and their constant monitoring and analysis by those making organizational policy. 8.25 Management information systems can be used to: i) diagnose and measure the nature and scope of the initial workload (e.g., numbers and types of crimes reported); ii) evaluate the performance and operational effectiveness of individual institutions within the system (e.g., numbers of police investigations completed and of cases transferred from police to prosecutors, percentage of investigations or tried cases resulting in a conviction, recidivism rates upon release from prison, etc.); iii) determine whether the individual institutions in the system (police, prosecutors, judicial system, prisons) are 112 working in alignment;142 and iv) identify the impact of the system on the broader objective of reducing crime and violence. They can also be used to develop a series of indicators to track the most critical outputs and impacts and to help outside observers understand and evaluate the reformers' efforts. 8.26 Indicators are best produced not by ad hoc efforts to measure outputs currently of interest, but through the routine registration of basic data on work processes by those carrying them out, and the transfer of key information to centralized databases where the data can be reviewed and analyzed for its broader implications. When the process is linked to the wide-spreading adoption of automated "case" files, it should not represent extra work for the front-line employees. Such systems, once in place, provide management a very potent tool for reviewing organizational performance and identifying and diagnosing problems. They also facilitate the creation of new indicators as they are needed. 8.27 The following sections draw on statistics currently recorded by the governments of the Dominican Republic and Jamaica to develop a series of comparative indicators on the performance of their criminal justice system. As the accompanying discussion suggests, the indicators are useful in signaling where problems may exist, but a full understanding of their significance and causes requires a contextual analysis of additional statistics and quite probably a review of the raw data on which they are based. MEASURING THE SYSTEM'S WORKLOAD: REPORTED OFFENSES 8.28 The first of all variables that must be tracked is the workload that the criminal justice system experiences. While the uniformed police in particular undertake a variety of tasks that does not involve the processing of individual cases, the number of reported offenses is the single best indicator of systemic workload. This information can be used in tandem with knowledge of the resources required to process an average case to get a sense of whether current resources are proportionate to the scale of the problem. 8.29 Having a unified set of national statistics on crime is critical to understanding organizational performance and improving crime prevention interventions--recognizing, of course, that under-reporting may mean that only a small fraction of certain crimes will be reported. Unfortunately, the government of the Dominican Republic is unable to generate a unified set of national figures on the number of crimes recorded. The collection of information about offenses in the Dominican Republic is left to the separate agencies of law enforcement (such as the National Division for the Control of Drugs) and, within the national police, divided up by different divisions responsible for their 142Systemic alignment allows the disparate elements of the criminal justice system to work together in harmony. It, for example, would allow: i) the prosecution to reinforce and multiply the effectiveness of the police, rather than inadvertently undermining police actions; ii) the provision of services to victims of past crimes to improve citizens' confidence in the criminal justice system and hence contribute to the reduction of future crimes; and iii) the appropriate sentencing of those committed of crimes (youth facilities, prisons, alternative sentences, etc.) to maximize rehabilitation. 113 investigation. Because these separate institutions use different language, frames, and reports to record these data, there is no clear national portrait of all crime.143 8.30 Jamaica, in contrast, aggregates information about all reports of crime in the department of statistics of the Jamaican Constabulary Force (JCF). The JCF receives reports from commanders in each parish and then sorts them into six categories of "major crimes." Unfortunately, at least according to one police official, the reliability of these figures is suspect.144 The Jamaican Constabulary Force had some 7,200 ranking personnel at the end of 2005 to deal with 1,674 murders, yielding a ratio of 4.30 officers per murder.145 For comparative purposes, in South Africa, there are some 120,000 police officers and 19,000 murders, for a ratio of six to one.146 In the Dominican Republic in 2005, there were about 26,000 police officers (Centro de Estudios de Justicia de las Américas, 2004-2005) and 2,400 murders in 2005, for a ratio of about 11 to one.147 While useful as a first approximation of workload, international comparisons of police/murder or police/population ratios are sometimes problematic.148 INSTITUTIONAL PERFORMANCE MEASURES Police Performance Indicators: Clearance Rates for Homicide 8.31 As in almost all countries, the police in Jamaica and the Dominican Republic routinely measure their "clearance" rate at which dockets leave their authority and become the primary responsibility of the prosecuting authority. 149 This is a key performance indicator for both uniformed personnel and investigating officers, as both parties play a key role in gathering information and witnesses, and apprehending suspects. Acceptance of the case by prosecutors and judicial authorities indicates that the police have built a sound case. Police do not generally track the processing of the case through the justice system or otherwise monitor the rates at which arrests yield 143 Within the national police, the function of analyzing crime and violence is greatly dispersed. Beneath the Office of the Chief of Police (Jefatura), there are a few data analysts as well as a Department of Statistics, along with a more recent creation, the Department of Mapping. The Division of Criminal Intelligence (DINTEL), formerly the Secret Service, has its own analytical unit and statistics division. The Division for the Investigation of Crimes (DICRIM) is supported by a Department for the Analysis and Processing of Information, but its subordinate operational investigative units, such as the Departments of Homicide, Robbery, and Auto Theft, also have their own data collectors and analysts. Each of these departments and operations, moreover, receives different quantities and types of information. 144 Interview with senior police official. 145 This is below the official allocation of roughly 8,000 positions, an indication of accompanying recruitment problems. 146 http://www.saps.gov.za 147 The Dominican government does compile national homicide figures, released through the office of the Attorney General. 148 Differing distributions of types of crime across countries, different policing models (preventive vs. reactive, beat patrol vs. car patrol, etc.) mean that comparing ratios across countries is problematic. More fundamentally, the number of offenses reported is endogenous: improved police-community relations will result in more incidents reported to police and lower personnel to offense ratios, but this may be the result of good performance rather than an indicator of insufficient personnel and associated poor performance. 149 The police consider a crime cleared if a person is arrested for the offense or if they conclude that the crime was committed by a person who subsequently dies. Some police forces also consider a crime cleared if they have named a suspect (usually in an arrest warrant) who has not yet been apprehended. 114 convictions, as this largely depends on the activity of prosecutors. Of course, the police continue to play a role right up to the time of conviction, and therefore it is important that the performance of the system as a whole be monitored by a higher authority. 8.32 The national police force in the Dominican Republic does not track clearance rates for homicide separately from other crimes against the person (e.g., robbery, assault). Nevertheless, unpublished information compiled by the police about the number of reported violent crimes for which the police had made an arrest shows that in the last four months of 2005 and first month of 2006, the police recorded 1,643 "crimes against the person" and made arrests in 33 to 40 percent of such cases. 8.33 Analysts in DICRIM, the Division for the Investigation of Crimes, are trying to improve this indicator, for in its present form it does not distinguish between old and new cases (e.g., one cannot tell what portion of the 111 cases solved in September came from the 83 cases pending in August). They also hope to separate homicide from assault and robbery, which they suspect have lower rates of clearance, but they have yet to overcome the technical and administrative hurdles to such an assessment. 8.34 A more discriminating measure of clearance rates can be composed with data collected by the Statistics Department of the Jamaican Constabulary Force. Table 8.2 shows how this Department sorts reported homicides into categories by "motive"--that is, the circumstances which a police investigator or statistician considers to have caused the act of violence.150 It records clearance rates ("C/up") for each type of newly reported homicide.151 These vary from a low of zero in cases of drug-related murders in 2005, to a high of 83 percent for "other criminal acts" in the first six months of 2006. Table 8.1: Homicide Clearance Rates in Jamaica by "Murder Motive," 2005 and 2006 (January 1-July 16) 2006 2005 C/up C/up Murder Motive Reported % of total N % Reported % of total N % Drug Related 15 2% 11 73.3% 15 2% 0 0.0% Gang Related 159 22% 44 27.7% 224 24% 61 27.2% Domestic 22 3% 14 63.6% 26 3% 18 69.2% Other Criminal Act 335 47% 278 83.0% 423 46% 260 61.5% Not Yet Established 182 26% 0 0.0% 240 26% 0 0.0% Total 713 100% 347 48.7% 928 100% 339 36.5% 8.35 As a measure of police performance, clearance rates are key. The job of detectives or judicial police is to build a prosecutable case against a suspect in each instance of serious crime they record. Of course, cases may emerge to be unfounded or 150 Mr. Barrett, the deputy director of the statistics department, explained that, in reviewing police reports of homicide and other crimes, the disparity between the reported facts of the case and the account of the investigator sometimes compels him to redefine the motive for the murder. Personal interview, July 17, 2006. 151A separate table tracks the number of murders from "previous years" that are cleared. 115 the interests of justice may not support a prosecution, so clearance rates will never be 100 percent. Furthermore, "clearances" achieved at the scene of the crime--through witness identification, direct police observation, or even the death of the suspect--are weaker indicators of performance, since they are inflated by good fortune and violent confrontations.152 But given that the share of these "easy" clearances is not likely to vary over time, police can still strive to better the rate at which they handle the more difficult cases.153 8.36 In order to give an assessment of how these clearance rates compare internationally that is fair to the police, additional information is required, including the number of active cases per homicide detective. In general, though, the overall homicide clearance rate in Jamaica is comparable to that in other countries with sound police forces and similar levels of economic development. External Accountability Measures: Complaints against the Police 8.37 Of course, high clearance rates cannot be the only indicator by which the police are judged. Excessive zeal as well as baser motives can lead police to run afoul of the law, and success in combating crime must be leavened with respect for human rights. For this reason, it is essential that the number and nature of complaints against the police be tracked. 8.38 Complaints against the police are a mixed indicator. On the one hand, they reveal sometimes shocking misconduct, and the media coverage given to these events can undermine confidence in the system. On the other hand, they indicate that the public has sufficient faith in the system to report these offenses. Changes in the number of offenses recorded, particularly in periods of transition, could be more indicative of an increase in citizen awareness and participation in the complaints system than of real changes in the incidence of misconduct. 8.39 In Jamaica, two separate bodies collect complaints from civilians against the police: a Department of Internal Affairs within the JCF and the Police Public Complaints Authority (PPCA) within the Ministry of Justice. There is little public information about the work of either of these entities. Still, there are signs that the government of Jamaica is trying to improve their effectiveness as accountability mechanisms. The National Security Strategy for Jamaica, promulgated by the Ministry of National Security in May 2006, recommends that, as part of the strategic review of the JCF, there be an assessment that would "examine legislative and higher level administrative structures, such as the Police Services Commission (PSC) and Police Public Complaints Authority (PPCA), in 152Some police forces, including the JCF, consider a crime cleared if the suspect is killed. In nearly one- quarter of the cases of robberies considered cleared by the JCF, the suspect had been killed by police officers or civilians. 153Additional refinements to the clearance rate could help the police promote equity in their policing and adherence to certain human rights standards. For example, police statisticians could develop a measure of equity by distinguishing between clearance rates in poor versus affluent districts--recognizing, of course, that factors beyond the police control in the short run such as problematic police-community relations may reduce clearance rates in poorer districts. 116 order to facilitate an effective civilian oversight mechanism for proper accountability for its performance and conduct," (Ministry of National Security, 2006). 8.40 In the Dominican Republic, the Office of Internal Affairs is charged with holding police accountable for misconduct. Its director is appointed by the President, not the Chief of Police, and thus enjoys some independence and autonomy. This office has an independent investigative capacity within its staff of 214, although its resources are stretched. It can initiate its own investigations as well as act on complaints it receives from the public. In the past, when this office recommended disciplinary proceedings against police officers, these were typically adjudicated by a police tribunal. Increasingly today, however, according to the new director, complaints substantiated by the office are brought before ordinary courts.154 8.41 Between August 2005 and May 2006, the period for which detailed information is available, the department received 637 complaints, an average of two per day.155 Nearly three-quarters of these complaints were substantiated through investigations, an extraordinarily high percentage for any country. Just over half of the complaints (319) involved allegations of physical aggression or death threats by officers. This indicator suggests both high levels of misconduct and good performance by the Office of Internal Affairs in holding individual police officers accountable to professional standards. 8.42 The office is currently designing a stronger information management system, which should improve both the reliability and the range of data available for performance measurement. If that new system were to include information (currently not recorded) about the complainants, it would allow the office to measure the equity in its own responsiveness, especially to poor and vulnerable groups. 8.43 In neither Jamaica nor the Dominican Republic do the formal accountability mechanisms reach beyond oversight of the police. There is no national justice commissioner or ombudsman for the receipt and investigation of complaints against other institutions in the criminal justice system. Prosecution and Adjudication Performance Indicators: Prisoners Awaiting Trial or Sentence 8.44 The most common indicator for measuring prosecutorial performance is the rate of convictions. While a certain share of cases accepted from the police may be dismissed in the interest of justice, the decision of the prosecutorial authority to undertake the expense of a trial must be hinged on a belief that the accused is guilty and the case can be 154Interview with General Daisy Liriano. 155In 2005, the department of internal affairs investigated nearly 7,000 matters of disciplinary and other infractions. It requested indictments for 123 officers, and submitted another 363 to the police tribunal to consider their suspension or removal from the national police force. In addition to this sum, 111 officers were suspended, and another 96 released from the force as a result of complaints of domestic violence, improper use of firearms, robbery, and other acts. See "Memoria Annual 2005," unpublished report, National Police. 117 won. The number and nature of convictions therefore remain the primary measure of prosecutorial effectiveness. 8.45 But in addition to effectiveness, the efficiency and the equity of the prosecutor's work are key. Trial verdicts must be delivered as quickly as possible so that the innocent may be released and the guilty can undergo rehabilitation. Failure to do so has serious consequences, including overcrowding of prisons. As a result, the share of prisoners awaiting trial is a another key indicator of criminal justice system performance. 8.46 In countries where the system has collapsed, such as Haiti, as many as 80 to 95 percent of all inmates may be awaiting trial, (Vera Institute of Justice , 2006; National Center for State Courts and USAID, 2006)156 and the situation is nearly as bad in many parts of the Caribbean. Undue delays in the processing of cases are a violation of the human rights of prisoners awaiting trial. In addition, in countries like the Dominican Republic, where prison cells are seriously overcrowded, rehabilitation work becomes impossible, and again, human rights concerns emerge. It is thus essential that criminal justice systems strive to reduce the share of prisoners awaiting trial (see Clouatre, 2002).157 8.47 Good management statistics can help identify the problems and their sources and track the impact of reform measures. However, few courts, and neither of those surveyed here, currently capture sufficient statistical data on system-wide performance to permit this kind of analysis158. Thus, we are left with one global indicator, the percentage of unsentenced prisoners. 156"Prolonged Pretrial Detention in Haiti", July 2002 gives the lower figure. A recent study by the National Center for State Courts indicates that in Port-au-Prince, the numbers in 2005 approximated 95 percent, although nearly all the detainees faced charges for major felonies. 157There are several ways of doing this. One is to increase the efficiency of the prosecutorial process, which involves coordinating a wide range of criminal justice actors. The police must take the capacity of the justice system into account when engaging in discretionary arrests, such as those encountered in mass "sweep" operations. Judicial hours must be reasonable. (In many countries, judges sit for absurdly short hours each day due to entrenched traditions that have not responded to present-day reality. In Haiti, for example, the courts of first instances (Courts of Assizes) conduct hearings on serious felonies only twice a year, thereby extending pre-trial detention periods Court management (record keeping, scheduling practices) is important. The provision of pre-trial screening and legal reforms can allow for the release of many non-violent accused who would otherwise be held with no bail or with bail set beyond their means. 158Where adequate records are kept at the courtroom level, analysts can use case file samples to investigate potential problems ­ delays and their causes, differential treatment of parties or types of cases, and even signs of corruption. However, absent direct access to web-based, electronic files, this is time consuming and costly work. For an example, see World Bank, 2003, An Analysis of Court Users and Uses in Two Latin American Countries. Washington D.C: The World Bank, Report No. 269666 118 Figure 8.1: Sentenced and Un-sentenced Prison Inmates in the Dominican Republic: May 2003 to May 2006 18000 16000 14000 12000 Prisoners 10000 of 8000 6000 Number Awaiting Appeal 4000 Sentenced Unsentenced 2000 0 May-03 Nov-03 May-04 Nov-04 May-05 Nov-05 May-06 8.48 In the Dominican Republic between May 2003 and September 2004, when the new code of criminal procedure went into effect, the total population in custody declined from 16,491 inmates to 12,806. As the Figure 8.1 shows, most of this reduction came from the ranks of unsentenced inmates and the number awaiting the outcome of appeals.159 From September 2004 through May 2006, however, there has been no major change in the number of inmates. The number of unsentenced prisoners initially declined, but then rose again toward the end of 2005; this group still comprises nearly 70 percent of all persons in custody.160 Prison Performance Indicators: Escapes, Deaths, and Violence in Custody 8.49 The role of the prison system is to prevent prisoners from committing crimes while in custody and to work toward their rehabilitation. The former goal is easily measured in the rate of escapes and the amount of violence in prisons. The latter, often summarized in some form of recidivism rate, is much more difficult to capture. 8.50 Figure 8.2 shows the indicators measured in the Dominican Republic, including deaths in custody and non-lethal violence. The indicators are difficult to interpret in this 159On average, another 100 to 140 inmates were in custody each day pending the ruling of a cassation court. 160Prisons in general are substantially over-crowded: the system has a legal capacity of just over 9,000. 119 form because they show absolute numbers rather than rates per prison population. As seen in the earlier discussion of pretrial detention, the prison population in the Dominican Republic was substantially higher in 2003 than the stable level it maintained from 2004 through mid-2006. When recalculated as rates, the level of deaths in custody appears to be steadily increasing from 48 deaths per 10,000 prisoners in 2003, to 58 in 2004, and to 68 in 2005. The figures for the first five months suggest that the prisons may be experiencing their first decline in recent years, back down to an annual rate of 49 deaths per 10,000 prisoners. The prison administration does not report the reasons and circumstances of these deaths, so the relative contribution of crowding, illness, and violence to these problems cannot be determined. 8.51 The changes in the number of escapes (Figure 8.2) are perplexing whether or not these are calculated as rates. The huge annual swings in the numbers of escapes reported make any comparisons with other countries meaningless, since the security of the prisons will seem relatively loose or tight depending on the year. In 2006, there were 24 escapes in the first five months of the year.161 Figure 8.2: Prison Deaths, Escapes, and Recaptures, Dominican Republic 100 90 88 80 77 75 70 60 62 58 50 50 40 30 27 20 22 16 16 10 14 12 0 2003 2004 2005 2006 (estimated) Deaths Escapes Recaptures Source: Dirección General de Prisiones, Procuraduría General de la República Notes: In 2005, an additional 138 inmates died during a fire. The estimate for 2006 is based on 26 deaths, 24 escapes, and 6 recaptures in the first five months of the year. 8.52 The Department of Correctional Services in Jamaica also tracks this kind of information with additional detail, including the level of non-lethal violence. The prisons in Jamaica report substantial reductions from five years ago in both the number of unnatural deaths and the number of violent incidents in custody (Figure 8.3). In 2001, 161The number of "recaptures" in 2003 exceeds the total number of escapes because of a large number of escapes in the preceding years. 120 there were 17 unnatural fatalities out of a total of 33 deaths in prison. In 2005, there were 8 unnatural deaths out of a total of 20. In 2000, there were 153 violent incidents reported in adult institutions. This number decreased to 92 in 2005. The numbers of violent incidents appears to have increased in 2004 and again in 2005, but the Director of the Department of Correctional Services attributes this to better systems of reporting. There were seven escapes from prison in 2004, and none the following year, although a "massive escape" from the Tower Street facility was apparently foiled in 2005.162 Figure 8.3: Unnatural Deaths and Violent Incidents in Jamaican Prisons, 2000-2005 200 180 185 160 153 140 120 100 105 92 80 60 54 40 35 20 17 8 6 8 2 2 0 2000 2001 2002 2003 2004 2005 violent incidents in custody unnatural deaths Source: Department of Correctional Services, Annual Report, 2005. 8.53 The Department of Correctional Services in Jamaica is eager to develop a wider range of indicators, including measures of recidivism. The Department is beginning to collect and report information on the number of persons admitted to prison who have previously been incarcerated. At first glance, an increase in this number might appear to reflect an increase in recidivism, but it is equally or more likely to reflect changes in the performance of other institutions of the system. For example, if increases in the numbers of police lead to increases in arrests, the numbers of prisoners with prior incarcerations should increase. If sentencing legislation is toughened, if prosecutors become more aggressive in their charging decisions or more effective at obtaining convictions, or if judges become more punitive in their sentencing, the result will be an increase in the number of prisoners with prior records of incarceration, even if recidivism remains flat. Perhaps most likely of all, if community surveillance of people returning from prison becomes more effective--as is occurring in almost every country--then returning prisoners will be especially vulnerable to re-arrest and re-incarceration, creating a false impression of rising recidivism. 162Some of this information is now available on the website of the Department of Correctional Services, http://www.dcsj.net/p/stats.htm 121 8.54 A better measure, and one more widely used globally, is the percentage of prisoners released in a particular year who are re-admitted within three or five years. This indicator is also ambiguous--responsive to enforcement efforts as well as changes in actual criminal conduct by released prisoners--but it is better than the alternatives. Moreover, it could be calculated by the Department of Corrections on its own, once it puts in place record systems that accurately identify new persons admitted with their prior records. 8.55 Even better, however, would be systems that track the re-arrest of released prisoners, or surveys that capture the experiences of released prisoners in much shorter time periods. Such data, focusing on recidivism in the first month, quarter, and year after release, would allow officials and prevention programs to identify some of the proximate causes of re-offending. IMPACT INDICATORS: USING CRIME AND ARREST DATA FOR CRIME PREVENTION 8.56 The discussion above has shown that the Dominican Republic and Jamaica, while making serious efforts to improve, still have a way to go in improving their performance measurement and the basic functioning of their criminal justice systems. The systematic collection of performance statistics, the generation of key indicators, and benchmarking for improvement should become a routine part of criminal justice system management, and performance against these indicators should be published and used as part of the budget allocation process. These figures should also be made available at the precinct level, so that local constituencies can hold their public servants accountable. 8.57 Could the criminal justice systems in these two countries be doing more to reduce crime and violence? That question necessitates looking beyond the indicators discussed above and delving more deeply into the mechanisms by which the activities of the justice sector actually contribute to reductions in crime and violence. In countries around the world, officials are asking themselves such questions, producing a global shift of strategic thinking within criminal justice. 8.58 More sophisticated indicators must capture the extent to which the criminal justice system acts proactively to eliminate the proximate causes of crime and violence. Such a criminal justice system would analyze crime patterns and use arrest powers strategically. Measuring and Understanding Crime: Analyzing Crime Patterns 8.59 Officials in the Dominican Republic are experimenting with ways of using the data that allow them to see more deeply into the proximate causes of crime that law enforcement might then target. Specifically, the police in the Dominican Republic are beginning to track the level, timing, and movement of crime as part of the Barrio Seguro Program, using digitalized maps and census data from the National Office of Statistics (see Box 8.1 for a description of the Barrio Seguro Program.) An initial draft of a "calendar of crime," prepared by the Unit for the Analysis and Processing of Information within DICRIM, records the time and days in which offenses committed in public in each 122 of the neighborhoods tend to occur (Secretary of the State of Interior and Police of the Dominican Republic).163 8.60 These efforts to measure and understand crime are promising not only because the police can adjust the frequency, routes, and number of officers on patrol in response to this analysis, thereby increasing the efficiency of deployments, but also because the public can use this knowledge both to avoid victimization and to more effectively participate in efforts to reduce crime. Measuring Impacts of Specific Criminal Justice Instruments: Reducing crime by using arrest powers strategically 8.61 Because the police often conceive of arrest as the end of their work, prosecutors and courts see it as the beginning, and the work of managers of jails and prisons can be burdened by unnecessary arrests, governments need impartial assessments of the system- level impact of arrest on crime and justice. 8.62 Based on data supplied by the police, it is possible to estimate that there currently are between 50,000 and 55,000 arrests each year in the Dominican Republic. Few of these roughly 50,000 annual arrests are pursued further by the justice system. According to data from the Ministerio Público, judges granted 12,237 requests to continue an arrest in the twelve month period of September 2004 to August 2005. In 5,318 of these instances (43.4 percent), a judge later ordered the suspect held in pretrial detention pending the outcome of trial.164 Beyond this, very little can be said as the government does not track the proportion of people arrested who were later held accountable for their behavior by the justice system. 8.63 The use of arrest powers by the police in the Dominican Republic, in short, is not typically followed by prosecution and even less likely to be followed by imprisonment. From a crime reduction perspective, this might be a good thing. If serious crime is reduced in areas where such stops are made, the police may be making an important contribution to public safety. Moreover, if the police are conducting stops within the law, and if the consequences for those stopped are brief and bearable, then the police may be using arrest powers in ways that reinforce the rule of law. However, to ensure this is the case, there need to be systems in place to verify that arrest powers are being used efficiently and within the law. 163These efforts are being supported by a partnership between John Jay College of Criminal Justice and the Secretary of the State of Interior and Police of the Dominican Republic. See the report on the website of the Ministry, www.seip.gov.do/2006/web/acuerdos_documentos.php 164The Supreme Court is in the process of installing a system by which to track the number of requests for arrest and detention that are granted by the judiciary. Personal Communication with Nestor Borroa, Department of Statistics, Supreme Court of the Dominican Republic. 123 Box 8.1: Using Crime Data and Analysis to Formulate a Multi-Sectoral Crime Prevention Strategy: the Barrio Seguro Program in Capotillo In July 2005, the government of the Dominican Republic launched the Democratic Security Plan (Plan de Seguridad Democratica). This plan was prepared by the Secretary of Interior and Police, the National Police and the Office of the Attorney General with the support of the Inter-American Institute of Human Rights, the John Jay College of New York, the University of Florida and Newlink Political Consulting firm. The Plan includes both preventive and control measures to reduce crime. The Plan established specific objectives based on the diagnosis of the security problem in the Dominican Republic. Based on a situational analysis and after identifying the cities/barrios with the highest level of crime, a pilot plan called Barrio Seguro (Safe Neighborhood) was implemented in Capotillo with the intention to both reduce and prevent crime. Capotillo, compared to seven other crime-affected neighborhoods in the National District, had recorded the highest levels of incidents related to drug trafficking, street delinquency, armed robbery, organized crime and gangs, as revealed by background analysis for the Democratic Security Plan. From January to August 2005 there were 30 homicides recorded in Capotillo; this is equivalent to an annualized homicide rate of 118 per 100,000 residents. The inability of the police to control crime in Capotillo had led to deterioration in public confidence: only 17% of the population believed that the police was doing a good job, and 80% of the population considered the police to be corrupt. Barrio Seguro was designed on the basis of data reported in the Democratic Security Plan. The program began in August 2005 and key interventions included: · Increased police patrols in 30 previously identified hot spots. · The addition of 230 specially trained police to patrol Capotillo. They were equipped with 3 new vehicles and 14 motorcycles for street patrol, as well as with a refurbished police station. · Situational crime prevention measures included infrastructure improvements to: i) facilitate access to and exit from the area; ii) provide street lighting; and iii) create new recreational areas for public use. · Investments targeting youth included: construction of new classrooms and development of cultural workshops and sports clinics. · General prevention programs included literacy programs and civic educational promoted by the office of the First Lady through its Progresando Program which is targeted to educate and help single mothers and their children. In total, the government allocated RD$46 million pesos (US$1.4 million) to execute the Plan in Capotillo. Barrio Seguro began to demonstrate positive results in Capotillo after only two months. Assaults and robbery declined 85 percent between early August and the end of September, 2005. No crime-related death was reported in this two-month period. From January to August 2006, there were 9 homicides in Capotillo, compared to 30 registered during the same period in 2005--a 70 percent reduction. The success of the Plan in Capotillo inspired its replication in 12 other neighborhoods, beginning in January 2006. While it is too early to definitively categorize Barrio Seguro as a success story (the project has not yet been subject to a rigorous impact evaluation), the approach is promising. It incorporates many principles that have shown to be effective internationally: i) complementarity between prevention and control initiatives; ii) policing focusing on hot spots and based on quantitative data on crime patterns; iii) use of situational crime prevention measures; and iv) involvement of the community and civil society organizations. 124 8.64 Specifically, two features of the arrests raise concerns. Only a small proportion is reviewed by prosecutors or scrutinized by courts, and the stops themselves are not subject to independent oversight. A modern justice system that relies on stops and arrests for minor offenses to suppress crime still needs a means to assure that the police are acting within the law in their interactions with the public. Equally important, the impact of these stops and arrests on crime is uncertain, for the police do not yet analyze whether their stops are followed by a reduction of crime in the immediate vicinity of the police actions. 8.65 In the case of Jamaica, it is possible to provide a superficial analysis of the relationship between arrest and reported crime. In 2005, according to police data, there were 15,746 arrests in Jamaica. In the same year, there were 8,461 reported "major" crimes--murder, shooting, rape, carnal abuse, robbery, burglary, larceny. There were, in other words, nearly twice as many arrests as major reported crimes. This ratio suggests that the Jamaican police are using arrest powers to deal with many "minor" offenses.165 8.66 In Jamaica, there is reason to worry that the use of arrest powers is not strategically targeted on high crime areas. The overwhelming majority of arrests in Jamaica took place in "rural" regions--that is, outside the metropolitan area where nearly half the population resides. The name "rural" may be misleading, since many rural areas have urban centers with populations over 100,000. Still, 72 percent of all arrests took place outside the metropolitan area, while 57 percent of all major crimes were reported inside the metropolitan area. 166 It is possible that recorded crime is lower outside the metropolitan area precisely because arrest powers are used so vigorously; alternatively, it is possible that crime is heavily under-reported in "rural" areas, or, as police now claim (see above) that urban criminals have been driven outside the major cities. Still, this anomaly calls out for further investigation. 8.67 A third way to measure the strategic use of arrest powers is to examine the relationship of arrests to imprisonment. This calculation can be made in both Jamaica and the Dominican Republic. In Jamaica in 2005, the average daily prison population was reported as 3,982, approximately one-fourth of the number of annual arrests. In the Dominican Republic that year, the prison population was 12,858, or roughly the same ratio to arrests as in Jamaica. Again, these ratios could indicate that arrests are being used to manage crime in many cases wholly apart from penal sanctions. This use of arrest powers to manage crime directly, rather than merely to take an offender the first step on the way to prison, is increasingly considered a sign of good police strategy if it is targeted, measured for its effect on crime, and reviewed by an independent authority (Weisburd and Braga). It is important that these steps be taken in the Dominican 165For example, state and local police departments in the United States in 2004 cleared 62.6 percent of murder offenses, 55.6 percent of aggravated assaults, 41.8 percent of forcible rapes, and 26.2 percent of robbery offenses. The data for major property crimes showed that agencies cleared 18.3 percent of larceny- theft offenses, 13.0 percent of motor vehicle theft offenses, and 12.9 percent of burglary offenses. 166The data on recorded crimes and police arrests comes from the Department of Statistics of the Jamaican Constabulary Force, which generously shared information and insights about crime analysis and problem identification. 125 Republic and Jamaica, especially in light of local critics' contention that such practices serve other, less desirable purposes.167 CONCLUSION: FROM ORGANIZATIONAL REFORM AND PERFORMANCE MEASUREMENT TO INTERAGENCY GOVERNANCE 8.68 In both case study countries and throughout the Caribbean, the on-going organizational reforms and stepped-up efforts to measure performance will require concerted attention over the near to medium term. However, to achieve their broader goals, the reforms will also require inter-agency governance mechanisms and incentives for individual agencies to align their individual programs. If sector-wide governance mechanisms are to succeed, their authority must go beyond performance monitoring to include real executive powers.168 The recent history of police and justice reforms in both Jamaica and the Dominican Republic illustrates this need. 8.69 Today, both governments are poised to introduce better instruments of interagency governance. In December 2005 and again in May 2006, the Ministry of National Security of Jamaica disseminated for public comment a green paper on public safety, in which it proposed, among other things, to locate responsibility for the implementation of such a strategy in the office of the Prime Minister (Ministry of National Security, 2006).169 In July 2006, the President of the Dominican Republic decreed the establishment of a National Security Council, whose job is to "plan, produce, coordinate, and evaluate the use of inter-institutional information for the prevention of crime."170 8.70 The plans for the Security Council in the Dominican Republic are more advanced than in Jamaica, and yet its success is not assured. In order to succeed, the new National Security Council will need a staff and other resources that make it a real instrument of governance, as well as a composition that does not simply represent the leadership of individual agencies or repeat membership on other inter-governmental committees. It will also need special skills to ensure the alignment of criminal justice system statistics and indicators and to coordinate efforts to prevent crime. 8.71 Finally, in both Jamaica and the Dominican Republic, innovations in governance should look beyond the criminal justice system for assistance in crime prevention. Victim services, indigent defense, addiction treatment, other public health services, and youth programs might all be brought together in the service of reducing crime and violence. 167See Bobea, op cit. As Bobea's article predates the new Dominican police law, which she helped to promote, there may have been positive changes since its enactment. However, considering long-standing, and well-documented abuses in that country, the introduction and careful monitoring of a more strategic and transparent approach to the use of arrests would seem especially essential. 168An example of sector-wide coordination currently limited to performance measurement is the role played by the National Commission for the Coordination of the Reforms in Chile. For a discussion of these challenges, see Measuring Progress Toward Safety and Justice: A Global Guide to the Design of Performance Indicators Across the Justice Sector, Vera Institute of Justice, 2004. www.vera.org/indicators 169The non-government organization, Jamaicans for Justice, released a parallel strategy paper, "Road Map to A Safe and Secure Jamaica," whose preparation was supported by an opposition leader. 170For a report of the decree, see www.presidencia.gov.do/frontend/articulo.php?id=4225 126 Governance structures need to be created to manage these efforts as part of a coordinated crime prevention strategy. 127 9. GUNS AND CRIME: A CASE STUDY OF TRINIDAD AND TOBAGO171 The rise of crime in the Caribbean has been characterized by the increased use of more powerful weapons, resulting in higher mortality levels. In 2004, Trinidad and Tobago experienced 160 firearm murders in 2004, 450 firearm woundings and 1500 non-injury firearm incidents. A major factor contributing to the surge of guns-related criminality in the region is the trafficking of narcotics which has facilitated the availability of firearms. The firearms required for protection of the contraband during transportation are smuggled in along with the drugs. Within these environments which promote the demand for weapons, reducing gun ownership is a difficult undertaking. Better gun registries, marking and tracking can help, as can improved gun interdiction in ports. In the long term progress will hinge on changes in the drug trade, changes in the gun culture, and progress in the implementation of international treaties and agreements on Small Arms and Light Weapons. 9.1. As violence has increased in the Caribbean, so too has the use of firearms. Increasingly, more powerful weapons are being used, resulting in higher mortality levels. The Caribbean has a long history of smuggling, and, as in the past, firearms are used in the transportation of illegal goods. "Guns and the illegal trade in drugs have formed a symbiotic relationship which has seen the emergence of increasing violence throughout the communities regionally" (WINAD, 2006). This chapter describes the situation in the Caribbean with respect to guns, crime, and violence. Trinidad and Tobago was selected as a case study, as it has seen a particular sharp rise in firearm-related violence over the last few years. 9.2. The CARICOM Regional Task Force on Crime and Security recently commissioned a report on the proliferation of small arms and light weapons (SALW) in the Caribbean (CARICOM, 2002). The resulting report identified three levels of SALW proliferation in the region: countries with established high levels and patterns of armed crime (Jamaica); countries with emerging high levels of armed and organized criminality (Guyana, Trinidad and Tobago); and countries with indications of increased use and availability of small arms (Antigua and Barbuda, Barbados, Dominica, Grenada, St. Kitts and Nevis, Saint Lucia, St. Vincent and the Grenadines). 9.3. At that time, it was determined that, among CARICOM countries, only Jamaica fell in the first category, with indications that military-type weapons were available and that paramilitary units were operating (Burrows and Matthias, 2003). If such an evaluation were done today, Trinidad and Tobago might also be included in this tier, as the murder rate has doubled since 2002. With 160 firearm murders in 2004, these were just the tip of the iceberg with above 450 firearm woundings and 1,500 non-injury firearm incidents. 172 171This chapter is based on a background paper prepared by Yvette Holder, consultant. 172Crime Reports from the Trinidad and Tobago Police Service. 128 Figure 9.1: Distribution of Injury Deaths by Type of Injury: Trinidad and Tobago, 1999-2003 300 250 hsta 200 de y 150 100 Injur 50 0 1999 2000 2001 2002 2003 Year Firearm Injuries Sharp Injuries Blunt Injuries Source: Central Statistical Office, Trinidad and Tobago. 9.4. For many countries of the English-speaking Caribbean, the majority of assaults and homicides were committed in past years with blunt or sharp weapons. The trend toward increased use of firearms in the commission of crime began in the 1970s in Jamaica, so that a court was established in 1974 solely dedicated to gun-related matters. In Trinidad and Tobago this change began in 2000. Before 2000, firearms were responsible for less than one-third of all homicides. By May 2006, this percentage had risen to 74 percent (see Figure 9.1 and Table 9.1). The percentage of homicides attributed to firearms in