combating crime and violent extremism using...
TRANSCRIPT
Combating Crime and Violent Extremismusing Forensic Economics
Quy-Toan Do
Policy Research Talk, 18 April 2016http://econ.worldbank.org/policyresearchtalks
Quy-Toan Do Economics of Crime PRT 2016 1 / 38
Forensic economics What is forensic economics?
What is forensic economics
The application of economic tools to the study of crime
Theory: criminals respond to incentives (Becker 1968)
“Supply” of crime/criminals“Demand” for crime = supply of opportunities to commit crime
Empirical investigations of market for crime (Ehrlich 1973)
Identification of supply versus demandPolicy: level of apprehension and punishment
Quy-Toan Do Economics of Crime PRT 2016 2 / 38
Forensic economics Crime and public economics
The “crime market”
Quantity
PriceD
S
Q̂
S: Incentives to commitcrime
D: Opportunities tocommit crime
Quy-Toan Do Economics of Crime PRT 2016 3 / 38
Forensic economics Crime and public economics
Combating crime is a public economics problem
Supply-side intervention:increase [opportunity] cost forcriminals
Quantity
PriceD
S
S’
Q∗Q̂
Demand-side intervention:decrease opportunities tocommit crime
Quantity
Price
D
D’
S
MSC
Q∗Q̂
Quy-Toan Do Economics of Crime PRT 2016 4 / 38
Forensic economics Crime and public economics
Measuring demand/supply price elasticity to guide policy
Impact of supply-side policies depend on (i) size of the curve shift,and (ii) slope of the demand curve
Quantity
Price D
D’
SS’
Q̂
P̂
Q̃+
P̃+
Q̃−
P̃−
Quy-Toan Do Economics of Crime PRT 2016 5 / 38
Forensic economics Crime and public economics
Three case studies
1 The African elephant poaching crisis
2 Piracy off the coast of Somalia
3 Violent extremists of Daesh
Quy-Toan Do Economics of Crime PRT 2016 6 / 38
The African elephant poaching crisis The magnitude of the problem
The African elephant poaching crisis
Quy-Toan Do Economics of Crime PRT 2016 7 / 38
The African elephant poaching crisis The magnitude of the problem
The African elephant poaching crisis
with
Julian Blanc (CITES)
Andrei Levchenko (U Michigan)
Lin Ma (National University Singapore)
Tom Milliken (TRAFFIC)
Quy-Toan Do Economics of Crime PRT 2016 8 / 38
The African elephant poaching crisis The magnitude of the problem
Poaching
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Africa
14606 carcasses
Est
imat
ed P
IKE
0.0
0.2
0.4
0.6
0.8
1.0
419-650,000 elephants inAfrica (2013)
3-5 million in early 20th
century
2011: estimated 7.4percent of elephantpopulation killed(monitored sites)
Quy-Toan Do Economics of Crime PRT 2016 9 / 38
The African elephant poaching crisis The magnitude of the problem
Quantity and prices
Price and quantities are two sides of the same coin
Price a necessary variable to estimate price elasticity of poaching
Law enforcement affects poaching through prices
Quy-Toan Do Economics of Crime PRT 2016 10 / 38
The African elephant poaching crisis The magnitude of the problem
Real elephant ivory prices (1970-2015)3
45
67
Ivory
price (
Chin
a−
HK
G; lo
g)
1970 1980 1990 2000 2010 2020year
CITES trade ban: 1989
Post-ban price growth:14% per annum
Ivory behaves like afinancial asset (store ofvalue)
Quy-Toan Do Economics of Crime PRT 2016 11 / 38
The African elephant poaching crisis The magnitude of the problem
Supply elasticity: poaching vs. ivory prices
2002
2003
2004
2005
2006
2007
20082009
2010
2011
2012
2013
20142015−
2−
1.5
−1
−.5
Avera
ge p
oachin
g index (
log)
4.5 5 5.5 6 6.5Ivory price (CHina−HKG, log)
Supply price inelastic: Elasticity ≈ .4
Quy-Toan Do Economics of Crime PRT 2016 12 / 38
The African elephant poaching crisis The magnitude of the problem
What did we learn?
Low supply price elasticity
price needs to drop a lot to bring poaching to sustainable levels
Decentralization of law enforcement and the role of communities
Ability and incentives to enforce lawFinancial benefits conditional on conservation: negative prices
Quy-Toan Do Economics of Crime PRT 2016 13 / 38
Pirates of Somalia
The pirates of Somalia
Quy-Toan Do Economics of Crime PRT 2016 14 / 38
Pirates of Somalia
The Pirates of Somalia
with
Jean-Baptiste Blanc (Shekere Ltd)
Aurelien Kruse (World Bank)
Trung Dang Le (Real Time Analytics)
Andrei Levchenko (U Michigan)
Lin Ma (National University Singapore)
Farley Mesko (Sayari Analytics)
Claudia Ruiz (World Bank)
Anja Shortland (King’s College)
Quy-Toan Do Economics of Crime PRT 2016 15 / 38
Pirates of Somalia Piracy in numbers
Somali piracy 2005 - ?
As of June 2015:
1,099 attacks
216 vessels hijacked
US$338m inransoms
4 years of captivity
Attack and Hijack Locations
AttackedHijacked
Legend
Quy-Toan Do Economics of Crime PRT 2016 16 / 38
Pirates of Somalia Piracy in numbers
Attacks and hijacks0
50
10
01
50
20
02
50
Num
ber
of A
ttacks
2005 2006 2007 2008 2009 2010 2011 2012
Year
Somalia Indonesia
West Africa Caribbean
Others
.1.1
5.2
.25
.3.3
5
Success R
ate
50
100
150
200
250
Num
ber
of A
ttacks
2005 2006 2007 2008 2009 2010 2011 2012
Year
Number of Attacks Success Rate
Quy-Toan Do Economics of Crime PRT 2016 17 / 38
Pirates of Somalia Piracy in numbers
Hijacking and ransoms
50
100
150
200
250
Ave
rag
e N
eg
otia
tio
n L
en
gth
(D
ays)
01
23
4
Ave
rag
e R
an
so
m (
Mil.
$)
2005 2006 2007 2008 2009 2010 2011 2012
Year
Average Ransom Average Negotiation Length
Quy-Toan Do Economics of Crime PRT 2016 18 / 38
Pirates of Somalia Offshore law enforcement
Modelling crime and deterrence on the high seas
Data limitations does not allow earlier methodology:
Alternative: take a modelliterally and choose coefficientsthat best explain:
pace of pirates improvedability
suddenness of attacks/fall
trends in ransom andlengths of captivity, etc.
2006 2007 2008 2009 2010 2011 20120
50
100
150
200
250
Year
Nu
mb
er
of
Att
acks
Data Model
2006 2007 2008 2009 2010 2011 2012−4
−2
0
2
4
6
Year
Ave
rag
e R
an
so
m
Data Model Confidence Interval
2006 2007 2008 2009 2010 2011 20120
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Year
Su
cce
ss R
ate
Data Model Confidence Interval
2006 2007 2008 2009 2010 2011 2012−0.2
0
0.2
0.4
0.6
0.8
Year
Ave
rag
e R
ela
tive
De
lay
Data Model Confidence Interval
Quy-Toan Do Economics of Crime PRT 2016 19 / 38
Pirates of Somalia Offshore law enforcement
Policy counterfactuals
Crime elasticity w.r.t. police
Optimal combination of navies/onboard security
Counterfactual 1:no onboard security
2006 2007 2008 2009 2010 2011 20120
50
100
150
200
250
Year
Num
ber
of A
ttacks
Data Model Counter−factual
Counterfactual 2:no navies
2006 2007 2008 2009 2010 2011 20120
50
100
150
200
250
300
350
Year
Num
ber
of A
ttacks
Data Model Counter−factual
Quy-Toan Do Economics of Crime PRT 2016 20 / 38
Pirates of Somalia Offshore law enforcement
Navies and onboard security guards
Security guards critical to explaincollapse of piracy in 2012
75% of large vessels to be equippedwith armed guards is sufficient
US$50,000 per trip = US$500mannually (or US$10+ bn in NPV)
Significant “pecuniary externality”Ransom totals approx US$50m/year
% Armed Guards
Social Costs MPBMPC
Quy-Toan Do Economics of Crime PRT 2016 21 / 38
Pirates of Somalia Rebuilding on-shore institutions
Using AIS to locate hijackedvessels
Mooring locations
Quy-Toan Do Economics of Crime PRT 2016 22 / 38
Pirates of Somalia Rebuilding on-shore institutions
Implication
Quy-Toan Do Economics of Crime PRT 2016 23 / 38
Pirates of Somalia Rebuilding on-shore institutions
And alas...
Quy-Toan Do Economics of Crime PRT 2016 24 / 38
The violent extremists of Daesh
Daesh – ISIL/ISIS
with
Mohamed Abdel-Jelil (World Bank)
Kimberly Baugh (U of Colorado)
Shanta Devarajan (World Bank)
Chris Elvidge (National Oceanographic and Atmospheric Admin.)
Jamie Hansen Lewis (Brown U)
Jake Shapiro (Princeton U)
Mikhail Zhizhin (NOAA & Russian Space Research Institute)
Quy-Toan Do Economics of Crime PRT 2016 25 / 38
The violent extremists of Daesh
Daesh – ISIL/ISIS
Islamic State in Iraq and the Levant
Origin in 1999 with Islamic State in Iraq (ISI): part of the Iraqiinsurgency against Western forces
Al-Nusra front in Syria following outbreak of civil war (2011)
Self-proclaimed ISIL/ISIS in April 2013
Quy-Toan Do Economics of Crime PRT 2016 26 / 38
The violent extremists of Daesh
A short history
Ibrahim Khalil
Al Yarubie
Abu Kamal
Azaz
Al Karama
Al Omari
Masnaa
Aboudiye
Attanf
Homs
Dar'a
Hamah
Idlib
Al-Kut
Samarra
An Najaf
Ba qubah
Al-Hillah
As-Samawah
Al 'Amarah
As Suwayda
Ad Diwaniyah
As-Sulaymaniyah
Mosul
Bayji
Duhok
Dayr az Zawr
Al Mayadin
Abu Kamal
Sa`dah
Basra
An-Nasiriyah
Irbid
ZarqaAzraq
Kirkuk
AleppoAr Raqqah
Karbala'
Ar-Ramadi Al Fallujah
Al-Hasakah
Shadadi
SinjarTall`Afar
Arbil
Aqaba
Latakia
Tartus
Tripoli
Um Qasr
BEIRUT
AMMAN
DAMASCUS
KUWAIT
BAGHDAD
KUWAIT
SYRIANARAB REPUBLIC
I R A Q
LEBANON
West Bank
Gaza
JORDAN
S A U D IA R A B I A
I S L A M I CR E P . O F
I R A N
T U R K E Y
ISRAEL
ARABREP. OFEGYPT
Homs
Dar'a
Hamah
Idlib
Al-Kut
Samarra
An Najaf
Ba qubah
Al-Hillah
As-Samawah
Al 'Amarah
As Suwayda
Ad Diwaniyah
As-Sulaymaniyah
Mosul
Bayji
Duhok
Dayr az Zawr
Al Mayadin
Abu Kamal
Sa`dah
Basra
An-Nasiriyah
Irbid
ZarqaAzraq
Kirkuk
AleppoAr Raqqah
Karbala'
Ar-Ramadi Al Fallujah
Al-Hasakah
Shadadi
SinjarTall`Afar
Arbil
Aqaba
Latakia
Tartus
Tripoli
Um Qasr
BEIRUT
AMMAN
DAMASCUS
KUWAIT
BAGHDAD
KUWAIT
SYRIANARAB REPUBLIC
I R A Q
LEBANON
West Bank
Gaza
JORDAN
S A U D IA R A B I A
I S L A M I CR E P . O F
I R A N
T U R K E Y
ISRAEL
ARABREP. OFEGYPT
Med
iter
rane
an S
ea
Euphrates
Tigris
ThartharLake
HabbānīyahLake
RazzazaLake
Euphrates
40°E 45°E
35°E 40°E 45°E
35°N
30°N
35°N
0 150 300
KILOMETERS
IBRD 42122 | MARCH 2016This map was produced by the Map Design Unit of The World Bank.The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.
GSDPMMap Design Unit
DAESH CONTROLLED FLARES
NON-DAESH CONTROLLED FLARES
MAXIMUM EXPANSION OF DAESHDURING PERIOD OF CONCERN
DAESH TERRITORY BY EARLYSEPTEMBER 2015
MAIN CITIES
NATIONAL CAPITALS
SECONDARY ROADS
PRIMARY ROADS
INTERNATIONAL BOUNDARIES
DAESH STUDY
Quy-Toan Do Economics of Crime PRT 2016 27 / 38
The violent extremists of Daesh How does Daesh finance itself?
Its financial sustainability
Oil
Smuggling and taxes
Looting and confiscation
Foreign donors
Kidnapping
Leakage of Iraq’s Central Bank foreign reserves
Quy-Toan Do Economics of Crime PRT 2016 28 / 38
The violent extremists of Daesh How does Daesh finance itself?
Oil revenues: what has been said?
Quy-Toan Do Economics of Crime PRT 2016 29 / 38
The violent extremists of Daesh Remote sensing
Estimating oil production
Quy-Toan Do Economics of Crime PRT 2016 30 / 38
The violent extremists of Daesh Remote sensing
The case of Ajil
Quy-Toan Do Economics of Crime PRT 2016 31 / 38
The violent extremists of Daesh Inference
Pre-war oil production
2015 2014
2013
2012
-20
24
6Lo
g(oi
l out
put)
-1 0 1 2 3 4Log(radiant heat)
95% confidence interval Point estimateIraqi fields Syria total
Beta=0.7999 Linearity test (p-value) = 0.1749 R-squared=0.4809
Quy-Toan Do Economics of Crime PRT 2016 32 / 38
The violent extremists of Daesh Inference
Daesh oil production
Quy-Toan Do Economics of Crime PRT 2016 33 / 38
The violent extremists of Daesh Inference
Revenues and production
We estimate an upper-bound of oil production that peaked at 31,000barrels per day in August 2014
Production then fell to a 2015 average of 18,000 barrels per day,yielding an approximate daily revenue of US$454,000
Quy-Toan Do Economics of Crime PRT 2016 34 / 38
The violent extremists of Daesh Inference
A similar supply/demand dichotomy
Dealing with the jihadistgroup’s finances
Oil is not (has neverbeen) a significant sourceof revenues
Extortion and taxes
External support
Fight radicalization in MENAcountries and beyond
Inclusive institutions andpolicies
Youth employment
Rebuilding the socialcontract
Quy-Toan Do Economics of Crime PRT 2016 35 / 38
Conclusion What did we learn?
Conclusion
What did we learn?
Quy-Toan Do Economics of Crime PRT 2016 36 / 38
Conclusion What did we learn?
International crime and violent extremism: thecommonalities
Crime and violent extremism: a labor intensive “sector”
Decentralization of law enforcement: capital/labor substitution andcommunity-driven enforcement
Ability of communities to enforce the lawIncentives for communities to enforce the law
Global spillovers
Mechanisms to solve collective action problemRole of multilateral institutions
Quy-Toan Do Economics of Crime PRT 2016 37 / 38
Conclusion On Forensic economics
Forensic economics and policy
Unconventional data collection for policy analyses of anunconventional economic activity
3 cases where evidence brings new perspective to policy consensus
Quy-Toan Do Economics of Crime PRT 2016 38 / 38