A Dynamic Analysis of Organized Crime in Jamaica
Jason Wilks’, Hugh Morris', Tameka Walker’, Matteo Pedercini’, Weishuang Qu?
1 Planning Institute of Jamaica, 10-16 Grenada Way, Kingston 5, P.O. Box 634, Jamaica, West
Indies
? Millennium Institute, 2200 Wilson Boulevard, Suite 650, Arlington, VA 22201-3357 USA
"tel: (+876) 906-4463-4, fax: (+876) 906-5011
* tel: (41-703) 841-0048, fax(+1-703) 841-0500
email: Jason_wilks @pioj.gov.jm, hugh_morris @pioj.giv.jm,
tameka_walker@pioj.gov.jm, mp @millennium-institute.org, wq@millennium-institute.org
ABSTRACT
The impact of the reform efforts of the 1990s, as well as the rapid process of globalization, all
resulted in far reaching changes in the structure of the Jamaican economy. With these changes,
the existing planning and economic models became increasingly limited in scope as they could
not adequately reflect the new structural reality. In March 2006, the Government of Jamaica
engaged the Millennium Institute to assist it in development of a modern planning tool with the
capability to integrate the economic, environmental and social elements. An important
component of the model is the sector for organized crime behaviour in Jamaica. The purpose of
this paper is to explain the development of the organized crime sector within the T-21 Jamaica
model and demonstrate the possible utility of system dynamics in facilitating discussions on
public policy.
Introduction
Jamaica, a land of natural beauty and heart-felt hospitality is known internationally on a scale
that belies its relatively small size. This is particularly the case, with its contributions to the
world of music and athletics. Another feature of the island known globally, that speaks more to
notoriety than fame, is the country’s battle with violent crime. Even while considering the
internal strife of Ireland, the drug wars of Colombia and the civil unrest of various South African
states, Jamaica held the dubious honour of “The Murder Capital of the World” in 2006 (Lalah
2006).
The statistics bear out this difficult truth. While the overall crime rate for Jamaica has been
downward trending for the last 10 years, the murder rate has risen sharply over that same time
period. Indeed, for many Jamaicans the murder rate is synonymous with the real crime rate and is
the best indicator of the crime problem for the island. The incidence of murders peaked in 2005
at 1,674 resulting in a murder rate of 63 per 100,000 persons, the highest rate in the world that
year (PIOJ 2006). Similar yearly totals have kept this small island of just 2.6 million inhabitants
at the top of the world rankings for murders for the past decade.
The Jamaican Government has made numerous attempts to stymie the crime problem. The
government pursued a crime-management approach to resolving the problem, from increasing
the share of government expenditure for national security to enacting stricter crime-fighting
legislation. In the last four years, the local law enforcement agency, The Jamaica Constabulary
Force (JCF), emphasized the need to include the citizenry as active participants in the fight
against crime and to this end, a number of community-based initiatives have been successfully
implemented. Notwithstanding, the murder rate remains dramatically high and the need for a
greater response to the crime problem persists.
The need for a solution to the crime problem occurs against a backdrop of a renewed drive for
national development. The Jamaican government recently engaged in the creation of a National
Development Plan with the explicit goal of achieving developed country status for the island by
2030. To assist in the development of the plan, a dynamic simulation model, Threshold 21
developed by the Millennium Institute (Qu 1995), was enlisted to provide long-term, integrated
projections on the outcomes of possible policy scenarios. Threshold 21’s main relevance was to
show country-specific interactions of important variables (Qu 2001) and this was deemed to be
of great importance in possibly understanding crime in the Jamaican context.
The purpose of this paper is to introduce the process of developing the crime sector in the T21
model and explore the consequences of crime reduction measures to national development. The
paper will first provide the historical context in which this problem evolved. Prior research
relevant to the issue will then be presented in a literature review. A discussion of the dynamic
hypothesis follows and possible scenarios the model provides will be highlighted to establish its
validity for realistic analysis. Finally, constraints of the model’s scope of analysis will be shared
and opportunities for further meaningful research will be addressed.
It must be noted from the onset that our intent is not to explain all the determinants of crime in
Jamaica. There are elements of crime which are irrational and unpredictable and therefore would
be impractical to model. With this in mind, we relegate our analysis to incentive-based,
organized violent behaviour which we believe can greatly assist government intervention against
crime.
Background
Historical Context
A proper understanding of the current crime situation requires an analysis of the historical
antecedents that have shaped it. The island’s first major flare-up of gun violence was during the
1980 general election. This period saw the deaths of over 800 civilians as well as 27 police
officers, as rival factions supporting the two major political parties clashed for control of
political turf. These rival groups and their leaders known as “dons”, originally supported by the
political parties, acted as enforcers of political will and ensured that citizens within respective
constituencies voted in support of their preferred candidates. In exchange, the political parties
allowed these enforcers to operate with impunity in these locales, in effect becoming the rule of
law in some instances. These locales became known as “Garrison communities’’(Stone 1981).
These communities were established and continue to exist for one reason alone, the need to
survive. As the Jamaican economy suffered setbacks in the mid 90s, the lack of financial
opportunities for the poor created a welcome environment for the proliferation of the
international narcotics trade. Jamaica is used as a main trans-shipment point between Latin and
North America in the illegal drugs trade (Moncrieffe 1998). Turf wars which were once political
have therefore become drug-related and where political parties once had influence on these rival
gangs, this influence is often only an illusion at best (Levy 1996).
Literature Review
A relevant body of literature on modelling crime related issues, and in particular on the
determinants and effects of criminal behaviour, formed the foundation for the crime sector of
T21 Jamaica. This literature encompasses research from the System Dynamics field, as well as
from the Sociology and Economics fields. To complement this research, information has been
collected locally from experts within the JCF, public sector and academia on the specific
characteristics of Jamaica’s crime issues, via ad-hoc seminars and individual consultations.
Determinants of criminal behaviour
Research on criminal behaviour has traditionally focused on individuals’ behaviour (Siegel
1998). In parallel, in the last three decades, an increasing number of studies have been carried
out on macro-level dynamics of criminal behaviour (Bursik and Grasmick 1993).
Among these studies, eight major theories seem to have emerged (Pratt 2001): Social
Disorganization Theory; Anomie/Strain Theory; Absolute deprivation/Conflict Theory; Relative
Deprivation/Inequality Theory; Routine Activities Theory; Rational Choice/Deterrence Theory;
Social Support/Altruism Theory; Sub-cultural Theory. Our model of organized crime is not a
clear cut example of one specific theory of criminal behaviour, but its distinctive elements can be
found in at least three of the theories above.
The information gathered from research done in Jamaica (World Bank 1997) pointed towards six
major elements: (1) poverty, (2) unemployment, (3) education, (4) population density, (5)
perceived gain from criminal activities, and (6) critical mass of criminals. Our theory of criminal
behaviour is therefore more oriented towards the social disorganization theory (Shaw and
McKay 1972). Elements such as poverty, unemployment and poor education not only seem to be
clear determinants of criminal behaviour in Jamaica from direct observation, but also appear as
strong predictors of crime in several empirical studies (Pratt 2001).
Nevertheless, our explanation of criminal behaviour also includes explanatory elements from the
Rational Choice Theory, such as the perceived gain from criminal activities (Becker 1968). The
perceived gain from criminal activities is also a determinant that has been recognized as major
determinants in other System Dynamics studies of organized crime in Italy (Raimondi 2001) and
on crime in Colombia (Hernandez and Dyner 2001).
Based on the information gathered from the field, we also included among the explanatory
variables the population density, in order to explain the higher crime rates in urban areas.
Similarly, the presence of a minimum critical mass of criminals seemed to be a necessary pre-
condition for organized crime to take place. These explanatory elements are more closely related
to the sub-cultural theory approach (Fisher 1975).
Effects of crime on economic performance
The effects of criminal activities on socio-economic development are of various natures:
monetary costs, opportunity costs, human capital costs, and other non-monetary costs (e.g. pain,
fear, etc.). Attempts to estimate these costs are particularly difficult in developing countries,
where data series on criminal activities are missing, or at best incomplete. Our model represents
only the monetary and opportunity costs of organized crime on economic development, and our
estimations are based on both cross-country studies and local data (Harriott et al. 2003).
Various studies have recently attempted to estimate the cost related to criminal activities.
Freeman (1996), provides generic figures for the cost of crime on GDP for the United States.
Bourguignon (1999) identifies various types of costs associated to crime, distinguishing between
industrialized and developing countries. Londofio and Guerrero (1998) estimate the costs of
criminal activities in South America. Raimondi (2001) also offers a System Dynamics
perspective on the various possible effects of crime on economic development in Italy.
Jamaica has a very peculiar economic and social structure, which can hardly be assimilated to
that of other countries in South America and Caribbean. Thus, the applicability of the findings of
the research above indicated to Jamaica is limited. Nevertheless, these studies assisted in our
estimation of the effect of organized criminal activities on Jamaica’s economic development.
Our own research, based on the collection of local data on criminal and economic activities and
the opinions of experts from the field, led us to the estimates used in the model.
Crime Modeling
Important Aspects
The first aspect of the research was to identify how the issue of the alarming crime rate was to be
studied. A superficial analysis of the official incidence of murder statistics for 2005 shows the
primary motives for murder as: (1) due to other criminals acts; (2) undetermined; and (3) gang
related activities; in order of prominence. Upon further enquiry, however, officers within the JCF
explained that in the majority of murders committed in the island, the motive was gang-related
turf war or reprisal killing. The inability to prove the involvement of gangs in most cases, due to
intimidation and fear of reprisal, leads to its placement as the third primary motive. The officers
also advised that analyzing murder in isolation would not be the best indication of the desired
trend of violent crime in Jamaica. They explained that with regards to the high incidence of
shootings, in the vast majority of these cases the intent was to murder. They surmised, therefore,
that a combination of these two indicators would yield a more accurate analysis of the Jamaican
context.
Despite the uncertainty surrounding motives for committing violent crime in Jamaica, one thing
has remained the same, the characteristics of the typical violent criminal. Data has shown that
over the last 10 years, the majority of violent crimes are committed by males in the 15-24 age
group. In 2005 for example, persons in this sub-category represented 43% of murder suspects
and 48% of shooting suspects. Similarly, it has been noted that male youths in the 20-24 age
cohort accounted for the largest share of murder and shooting victims, respectively in 2005
(PIOJ 2006).
The derivative from these trends is that a key element in the fight against crime is to understand
the process by which a young man, living in a depressed community enters into a gang and the
possible paths his life may take. An important part of the conceptualization was determining the
unit of analysis. The question was “Should we focus on the gang as the unit of analysis or the
individual member of the gang?” The risk in examining either unit was the possibility of
negatively stereotyping certain communities as breeding grounds for crime. After consultations
with local experts on the matter it was decided that the case for intervention would be best made
by analyzing the individual and how macro-level determinants may dictate their behaviour. The
variables established as determinants at the macro level, and the nature of their relationships to
the crime rate, are exhibited below.
Dynamic Hypothesis in Crime Sector
Data on the murder and shooting indicators for the reference period were ably supplied by the
Statistics Department of the JCF. The formulation of a dynamic hypothesis for our key issues
was then undertaken in the crime sector as shown in Figure 1.
real pe
-<adult literacy rate> pe
a CONMENTY WORK TO
<TOTALLAND mt tat effect of preceived aE DUCECONVERIING
REN mili ‘economic gain on TIMESERIES
comerting
GINCONTROLTO
ie:
REDUCE ACCESSIBILITY
10 GANG aS
population density effect of social —_effact of education
a nent effect of etc muss
total population ane. | onconverting ted
* non convicted gang
effect of unemployment on { Ce aces members death
epotentiad «= emg ~~ ON rae Cecil 5p |7 {
unemployment rate> LS gang rebted murders
‘Non Gang —J Non Convicted Gang |" St stnotings
>} Members Svp | Members
ret change in xp Sinerting to gang T
sop of me youths 3 cwverage case
poverty in urban centers resolving rate>-———Pfconvicton
: AVERAGE tna
PROPORTION IN “wasp CONVICTION TIME eave AVERAG:
CANGAREAS TINE // gue youths 4 MORE ALITY RATE
SERIES y FOR CONVICTED
—_[-cormicted Gane,
weresloe \ tig ae |
population below poverty tm convicted gang:
in> <Populaion> nnemrbers deaths
probability of being
(sop stands fr socal vulnerable population) rehibiliated>
Figure 1: The crime sector in T21
In order to develop our dynamic hypothesis, we first identified the key stock variables to be
represented. As mentioned above, we decided to analyze criminal behavior by first studying the
subjects that actually perpetrate gang-related crimes. These subjects are often part of low income
families, and we thus considered male youths from low income families as the pool from which
gang members can be recruited. Male youths in this category (socially vulnerable population, or
SVP) can thus be either active (non convicted) gang members, convicted gang members, or non
gang members. Non Gang Members Svp, Non Convicted Gang Members, and Convicted Gang
Members have thus been represented as stock variables.
The variables in blue, such as total population, are input variables to the crime sector from other
sectors in T21. The two green variables on top right of the sketch, COMMUNITY WORK TO
REDUCE CONVERTING TIME SERIES and GUN CONTROL TO REDUCE ACCESSIBILITY
TO GANG MEMBERS, are exogenous variables representing policy interventions. The red
variable on the right, gang related murders and shootings, represents variables created in the
crime modules that will be used in other modules and is the major indicator to represent the
crime rate.
The first stock variable, Non Gang Members Svp, is the group of male youths in the age cohort
15 to 24, living in poverty within known garrison communities. This group may be at risk in
engaging in organized crime activities and becoming gang members. The rate variable
converting to gang, which leads from this group to the next group, Non Convicted Gang
Members, is determined by a combination of factors, namely unemployment rate, population
density, education, possible economic gain by being a gang member, gang critical mass, and the
policy intervention of community work to reduce the conversion rate from an SVP to a gang
member.
Non Convicted Gang Members increases with new recruits each year and recidivists and
decreases due to deaths or convictions. The total membership within this population is also the
key determinant of the major output from this sector, gang related murders and shootings. Not
all gang members are assumed to be actual murderers, as some members may play strictly an
advisory role or be involved in other illegal endeavours such as car theft or drug-trafficking.
The other variable that determines the output is the gun accessibility. Although we have no solid
data to measure this accessibility, local experts in the field think that it has been increasing since
1990.
The third stock variable of interest is the group of Convicted Gang Members. It increases with
convictions and decreases with releases (either becoming a non-gang SVP with rehabilitation or
returning to the gang as represented by recidivism) or deaths. The conviction rate and the
rehabilitation rate are further determined by the average case resolving rate and probability of
being rehabilitated respectively. These two last variables are modeled in the diagram Figure 2.
INITIALCASE
RESOLVINGRATE
AVERAGE CRIMES
INITIALGANG MMITTED BY EACH
RELATED MURDERS, ‘CONVICTION
<gang related murders murders and shootings —————t _“tvenige case
and shootings> eee Dee
amare sehtne pe ee
s er :
<u eee ents
population> ow
fo: pe security,
expenditure
INITIAL PROBABILITY
; OF BEING
<real defense REHABILITATED
aes Gang Spa herpes probability of being
mmbers> : lubilta
epee $m rehuive perprtoner, Plated
prison population security expenditure
SHAREOF GANG SAHITe SECURITY
MEMBERS IN PRISON Eee EXPENDITUREEFFECT
"TIMESERIES ONREHABILITATION
Figure 2: Average case resolving rate and probability of being rehabilitated
The variable average case resolving rate is assumed to be dependent on the expenditure given to
the security system measured on per capita basis, while the probability of being rehabilitated is
assumed to depend on the per prisoner security expenditure, as shown in Figure 2.
Crime impacts on other elements of the society. In T21 Jamaica, the relative change since 1990
in the incidence of gang related murders and shootings, is used to estimate the extent of crime’s
impact on some of the most important elements for Jamaica’s national development. These
elements and their derivatives are illustrated in Figure 3 below.
average case resolving rate conviction
crime effect on electricity demand — total electricity demand in kwh
fail gdp ratio
foreign direct investment —
private investment
relative gang related murders and shootings‘ consumption
private co private domestic savings
Private savings
total tourism incomes
‘yearly number of incoming tourists <
tourist water demand
Figure 3: Crime’s effect on other sectors in T21 Jamaica
Crime’s effect on these variables is not always intuitive. The effect of crime on electricity
demand is a positive one, for example, when crime increases persons increase their use of
lighting and electronics for security purposes. Another positive effect is seen on private
consumption. Foreign direct investment, average case resolving rate and the yearly number of
incoming tourists are assumed to be negatively affected by in the crime rate.
Crime Scenarios in T21
In this section we present the simulation results from four alternative scenarios: a business as
usual scenario; a scenario with higher investment in police force; a scenario with higher
investment in rehabilitation services; and a scenario with higher investment in poverty reduction.
Baseline
With the crime sector added to T21 Jamaica, it is possible to generate the baseline (business as
usual) scenario to fit to historical data and project the trends for the future. The “business as
usual” scenario means that the present type and intensity of government intervention will remain
unchanged over the projected time period. This includes the share of expenditure allocated to
national security and enforcement efforts to mitigate accessibility to illegal guns.
In Figures 4, the baseline results from T21 (blue line) are compared to historical data (red line)
for gang related murders and shootings. The fit of our baseline scenario with historical data is
quite good, when considering that the model is the simplification of a reality that is yet to be
fully understood. In 2005, for example, our model estimated gang related murders and shootings
to be to 2,040 while historical records showed 2,324.
gang related murders and shootings
4,000
3,000.
2,000
1,000
0
1990, 2000 2010 2020 2030
Tine (Year)
gang related murders and shootings : base. —————————————_ crrine/Year
gang related murders and shootings : JamData. ——————————_ crine/Year
Figure 4: Gang related murders and shootings
The declining trend beyond 2020 in gang related murders and shootings is due primarily to the
decline in the number of male youths and to a decline in poverty as shown in Figures 5 and 6,
respectively. As these two variables decline over time, there are fewer people within the SVP
who can potentially become gang members. This reduces the stock of active gang members and
likewise their key output of gang related murders and shootings.
10
400,000
350,000
300,000
250,000
200,000
male youths
eet TPE
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
‘Tine (Year)
male youths : history¥JanData Person,
male youths : historwWbase Person
Figure 5: Male Youths aged 15 to 24
average share of population below poverty line
04
0.3
0.2
0.1
prison population
6,000
5,000
4,000
3,000
2,000.
1990 2000 2010 2020 2030
Tine (Year)
prison population : base Person
prison population : JamData Person.
Figure 7: Prison population
It is also possible to use the model to estimate variables for which we do not have data. This can
facilitate discussions with professionals in the field to advance our knowledge in crime related
matters and test our measures in dealing with them. Figure 8 displays the trends of two such
variables: Non Convicted Gang Members and Convicted Gang Members
Non Convicted and Convicted Gang Members
1,500
0
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
Convicted Gang Members : base Person
Non Convicted Gang Members : base Person
12
Figure 8: Convicted and non convicted gang members
Crime Scenarios
As the crime sector is dynamically linked to the other sectors within the T21 model, such as
population, education, employment, poverty, and government expenditures, itss output, gang
related murders and shootings, will vary considerably with different development policies and
assumptions. Such policies and assumptions cover a wide range of areas, including government
fiscal policies, budget policies, future financial flows of grants, borrowings, remittance, and
foreign direct investments, private consumptions and savings, international oil price, natural
disaster, HIV infection rate, and government poverty reduction measures. For instance, in the
scenario named Investment, where private marginal saving rate and foreign direct investment
both increase 20% by the year 2030 over the baseline, there will be higher GDP, higher
government revenues, higher employment, less poverty, and less crime, of which the last is
shown in Figure 9.
gang related murders and shootings
4,000
3,250
2,500,
1,750
1,000
1990 2000 2010 2020 2030
Tie (Year)
gang related murders and shootings : base. ———————. crime/Year
gang related murders and shootings : Investment ————————————_ crrie/Year
Figure 9: Scenario comparison on gang crimes
Scenario 1: Crime Reduction
When government takes effective measures in communities to slow the rate of recruitment to
gangs and tighten gun control so that gang members will have less access to weapons, it will also
affect all other sectors. The Crime Reduction scenario assumes that both the conversion to gang
13
and the access to guns will be reduced by 20% by 2030. Figures 10 to 14 show the
improvements in GDP, employment, and poverty of this scenario over the base case.
real gdp at market prices
600 B
450B
300 B
150B
10)
1990, 2000 2010 2020 2030
Tire (Year)
real gdp at market prices : base Jmd96/Year
real gdp at market prices : CrimeReduction. ———————————__ Jn 96/Year
Figure 10: Crime Reduction Scenario Comparison on Real GDP
total employment
2M
17M
14M
800,000
1990 2000 2010 2020 2030
Tire (Year)
total employment : base Person
total employment : CrimeReduction Person
Figure 11: Crime Reduction Scenario Comparison on Total Employment
14
average share of population below poverty line
04
0.3
0.2
0.1
0
1990 2000 2010 2020 2030
Tine (Year)
average share of population below poverty line : base ————————————___ Dn
average share of population below poverty line : CrimeReducton ———————_ Dn
Figure 12: Crime Reduction Scenario Comparison on Poverty
4,000
3,250
2,500
1,750
1,000
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
gang related murders and shootings : history\base cime/ Year
gang related murders and shootings : CrimeReduction ctime/Year
Figure 13: Crime Reduction Scenario
Comparison on Gang Related Murders and Shootings
15
prison population.
6,000
5,000
4,000
3,000
2,000
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
Prison population : histor-Wbase Person
prison population : CrimeReduction Person
Figure 14: Crime Reduction Scenario Comparison on Prison Population
Scenario 2: Security
The government may take another direction in fighting crime and decide that they will increase
the share of expenditure to national security. In the Security Scenario, the share of expenditure to
national security is doubled by 2030. Figures 15 to 19 illustrate the impact of this policy on real
GDP, employment and poverty.
16
real gdp at market prices
600 B
450 B
300B
10)
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
real gdp at market prices : history\base Jmd96/Year
real gdp at market prices : security Jm96/Year
Figure 15: Security Scenario Comparison on GDP
total employment
2M
17M
14M
11M
800,000
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
total employrrent : history\base Person
total employnent : security Person
Figure 16: Security Scenario Comparison on Total Employment
average share of population below poverty line
04
0.3
0.2
0.1
(0)
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
on below poverty line : history\base Dmal
on below poverty line : security Dmil
Figure 17: Security Scenario Comparison on Poverty
4,000
3,000
2,000
1,000
10)
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
gang related murders and shootings : history\base cime/ Year
gang related murders and shootings : security ctime/Year
Figure 18: Security Scenario Comparison on Gang Related Murders and Shootings
prison population.
10,000
0
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
Prison population : histor-Wbase Person
Prison population : security Person
Figure 19: Security Scenario Comparison on Prison Population
Scenario 3: Welfare
A final scenario that the model provides is an illustration of poverty alleviation measures’ impact
on the situation. Bearing in mind that gang members come from poor families, government may
try to reduce their numbers indirectly by increasing welfare expenditure to poor families. In the
Welfare scenario, government increases welfare expenditure from 2.9 billion real Jamaican
dollars to 5 billion real Jamaican dollars and increasing the share of welfare to the lowest quintile
from 0.5 to 0.8. Figures 20 to 24 show the impact of this intervention on real GDP, employment,
poverty, gang related murders and shootings and the prison population as against the base case.
19
real gdp at market prices
600 B
500B
400 B
300B
200 B
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
real gdp at market prices : history\base Jmd96/Year
real gdp at market prices : Welfare Jm96/Year
Figure 20: Welfare Scenario Comparison on GDP
total employnent
2M
17M
14M
11M
800,000,
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
‘Time (Year)
total employrrent : history\base Person
total employrent : Welfare Person
Figure 21: Welfare Scenario Comparison on Total Employment
20
average share of population below poverty line
04
0.3
0.2
0.1
(0)
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
on below poverty line : history\base Dmal
on below poverty line : Welfire Dmil
Figure 22: Welfare Scenario Comparison on Poverty
4,000
3,250
2,500
1,750
1,000
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
gang related murders and shootings : history\base cime/ Year
‘gang related murders and shootings : Welfare cime/ Year
Figure 23: Welfare Scenario Comparison on Gang Related Murders and Shootings
21
prison population.
6,000
5,000
4,000
3,000
2,000
1990 1994 1998 2002 2006 2010 2014 2018 2022 2026 2030
Tine (Year)
Prison population : historwWbase Person
Prison population : Welfare Person
Figure 24: Welfare Scenario Comparison on Prison Population
Conclusions
All scenarios reflect an improvement when compared with the base case. This indicates that
government has a range of policy levers by which it can cause abatement in the crime rate. The
Welfare scenario, however, seems to be the least effective of the three. For all the selected
indicators, with the exception of gang related murders and shootings, the difference between the
intervention and the base run was negligible. These results imply that if government wants to
address crime it must do so using direct means as much as possible.
This leads us to the next two interventions, the crime reduction scenario and the security
scenario. Their effects on real GDP, employment and poverty were almost identical. Their
differences lie in their respective impacts on gang related murders and shootings and prison
population. We find that the Crime Reduction scenario slowed the projected increase in the
incidence of murders and shootings faster than the Security scenario and also showed a decline
in the incidence by 2015 as opposed to 2017 with the Security scenario. The Security scenario on
the other hand, showed a sharper decline in the incidence of murders and shootings, resulting in a
lower incidence of 1,291 than the Crime Reduction scenario of 1,598 by 2030.
22
An undesired consequence of the Security scenario is, with the increase in real defense
expenditure and the subsequent increase in the average case resolving rate, our prison population
will grow dramatically from 5,554 to 8,874. The Crime Reduction scenario would, on the other
hand, only see an increase of 100 prisoners over the base case. Based on our <
sumptions, any
attempt to increase the efficiency of our justice system should also be coupled with expenditure
to increase capacity of the prison system.
It should be noted that the model results are not predictions, but scenarios that help participatory
discussions among all stakeholders and enable decision-makers to better understand the linkages
among all areas of development. The results presented above are based on a preliminary version
of the model that will be refined over time.
Further Research
The research has a number of limitations. Historical data is generally unavailable, and the ones
used to calibrate the model have not been externally vetted and could be a source of error. The
model also does not seek to address the demand side dynamics for criminality such as the drug-
trafficking link. The model also has room for further refinement, such as total land area could be
disaggregated to rural and urban land areas to give a more accurate picture of the concepts they
portray. The two points of intervention as well as the perceived economic gains from crime could
also be elaborated on by expanding to separate modules to show the dynamics that determine
their behaviour.
The reality is however that even if some important parameters need to be further calibrated it
would still be a useful tool for national security policy makers in the island. Discussions on the
model’s result with such persons have confirmed this as it allows them to view the results of
certain key policy assumptions in a transparent and objective manner. The model highlights
those areas of the national security and justice system which are in dire need of attention. It could
also point to the data collection mechanisms that need to be implemented or to pre-existing
mechanisms which may have become outdated. Finally, the model represents an ideal starting
place to engage various elements of the society with a frank discussion on the perceived ills and
possible solutions to the crime problem.
23
References
Becker, G.S. 1968. “Crime and Punishment: An Economic Approach.” The Journal of Political
Economy 76:169-217.
Bourguignon, F. 1999. Crime violence and inequitable development. Washington, D.C.: World
Bank
Bursik, R.J. and H.G. Grasmick. 1993. Neighborhoods and Crime: The Dimensions of Effective
Community Control. New York: Macmillan.
Fisher, C.S. 1975. Toward a Subcultural Theory of Urbanism. American Journal of Sociology
80:1319-41.
Freeman, R. 1996. Why do so many young American men commit crimes and what might we do
about it?. Journal of Economic Perspectives 10 (1):25-42
Harrriott, A., A. Francis, C. Kirton, and G. Gibbison. 2003. Crime and Development: The
Jamaican experience. Kingston: University of the West Indies
Hernandez, J.J. and I. R. Dyner. 2001. Crisis in Colombian prison: cause or consequence of a
flawed judicial system. Paper presented at the 2001 International System Dynamics Conference,
July 23-27, in Atlanta, USA
Lalah R. 2006. Three double murders within nine hours. Jamaica Gleaner January 5, 2006
edition. http://Avww.jamaica-gleaner.com/gleaner/20060105/lead/lead5.html (accessed March
19, 2007)
Levy H. 1996. They Cry ‘Respect’! Urban Violence and Poverty in Jamaica. Kingston. Centre
for Population, Community and Social Change.
Londojio, J-L and R. Guerrero. 1999. Violencia en America Latina: epidemologia y costos, IDB,
Washington, Working paper R-375.
Moncrieffe, D. 1998. Gang Study: The Jamaican Crime Scene. Kingston: Ministry of National
Security and Justice, Criminal Justice Research Unit. Jamaica
Pratt, T.C. .2001. Assessing the relative effects of macro-level predictors of crime: a meta-
analysis. PhD dissertation, University of Cincinnati.
Planning Institute of Jamaica. 2006. Economic and Social Survey Jamaica 2005. Kingston:
Planning Institute of Jamaica
24
Raimondi, V. 2001. Organized crime and economic growth: a System Dynamics approach to a
socio-economic issue. Paper presented at the 2001 International System Dynamics Conference.
July 23-27, in Atlanta, USA
Siegel, LJ. 1998. Criminology. Sixth edition. Belmont, CA: Wadsworth
Shaw, C.R. and H.D. McKay. 1972. Juvenile Delinquency and Urban Areas. Chicago: University
of Chicago Press.
Stone, C. 1981. Democracy and Clientelism in Jamaica. New Jersey: Transaction Books.
World Bank.1997. Violence and Urban Poverty in Jamaica: Breaking the Cycle-Washington,
D.C. World Bank
Qu, W and G. Barney. 1995. The Threshold 21 Sustainable Development Model. Paper
presented at the 1995 International Systems Dynamics Conference. July 30-August 4, in Tokyo,
Japan.
Qu, W and G. Barney. 2001. T21 China: An Initial Application and Analysis. Paper presented at
the 2001 International Systems Dynamics Conference. July 23-27, Atlanta, USA.
25
