Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more ➡
Download
Standard view
Full view
of .
Add note
Save to My Library
Sync to mobile
Look up keyword
Like this
2Activity
×
0 of .
Results for:
No results containing your search query
P. 1
Pattern in Escalations in Insurgent and Terrorist Activity

Pattern in Escalations in Insurgent and Terrorist Activity

Ratings: (0)|Views: 1,459|Likes:
Published by Justin Martinez

More info:

Published by: Justin Martinez on Jul 01, 2011
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See More
See less

05/27/2012

pdf

text

original

 
Pattern in Escalations in Insurgent and Terrorist Activityhttp://www.sciencemag.org/content/333/6038/81.full[01.07.2011 18:45:50]
Home>
et al.
,
333
(6038): 81-84
ADVERTISEMENTADVERTISEMENTTo Advertise Find Products
+
Science 
1 July 2011:Vol. 333 no. 6038 pp. 81-84DOI: 10.1126/science.1205068
REPORT
Pattern in Escalations in Insurgent and Terrorist Activity
1
2
,
3
4
5
1
5
6
7
,
8
Author Affiliations
ABSTRACT
In military planning, it is important to be able to estimate not only the number of fatalitiesbut how often attacks that result in fatalities will take place. We uncovered a simpledynamical pattern that may be used to estimate the escalation rate and timing of fatalattacks. The time difference between fatal attacks by insurgent groups within individualprovinces in both Afghanistan and Iraq, and by terrorist groups operating worldwide, gives apotent indicator of the later pace of lethal activity.
In 1948, Lewis Fry Richardson found that the number of wars with a given number of fatalities follows anapproximate power-law statistical distribution as a function of the number of fatalities (
1
). Recent research hasshown that a similar statistical distribution arises for the number of fatalities in individual clashes and acts ofterrorism (
, 
), whereas the relative stability of these distributions over time allows an estimate to be given ofthe severity of future wars or clashes within an ongoing war (
 – 
). However, these existing studies saynothing about the operationally relevant questions (
 – 
)of how the underlying arms race evolves over time,or when fatal attacks might occur. Here, we confront these questions using fatality data obtained on anunprecedented daily scale from the ongoing insurgent war in Afghanistan and the recent one in Iraq. Our dataanalysis is freely available for public scrutiny: The coalition military fatality data come from the public Web sitewww.icasualties.org. Our analysis was done with the free downloadable tool Open Office, which runs on anycomputer platform. The supporting online material (SOM) for this paper contains step-by-step instructionstogether with Open Office worksheets. For Afghanistan, we include fatalities from the start of OperationEnduring Freedom in 2001 until summer 2010, when General Petraeus became commander of theInternational Security Assistance Force and the U.S.-led surge started. For Operation Iraqi Freedom, the datainclude fatalities from 2003 until summer 2010, when U.S. military action officially ended. The terrorism resultsare derived from a recent study, which used the database of the Memorial Institute for the Prevention ofTerrorism, comprising the 3143 fatal attacks carried out by the 381 known terrorist groups operating within theperiod 1968–2008 (
). Suicide bombing data for Hezbollah (1982–1985) and Pakistan militants (1995–2008)comes from the public Web sitehttp://cpost.uchicago.edu.For a wide range of human activities, the time taken to complete a given challenging task decreases withsuccessive repetitions, following an approximate power-law progress curve (
 – 
). This inspires us to analyzethe insurgents’ completion of fatal attacks against coalition military forces in a similar way (
).We calculated the best-fit power-law progress curve
, where τ
represents the interval betweenthe (
– 1)’th and
’th fatal day (one in which the insurgent activity produces at least one coalition militarydeath), where
= 1, 2, 3, etc.,
indicates the escalation rate. τ
1
is the time interval between the first 2 dayswith coalition military fatalities. (For global terrorism, a fatal day is one in which a particular terrorist groupproduced at least one death anywhere in the world.)
 shows the best-fit values τ
1
and for eachprovince in Afghanistan, for all forms of hostile death. The average number of fatalities per fatal attack is fairlyconstant in a conflict (
, 
, 
), hence it is in insurgents’ interest (or that of a particular terrorist group) that thetime between fatal days decrease rapidly, and hence is large and positive, while the opposite is true for the
military (or counterterrorism force). The scatter in τ
in
 is typical for real-world tasks (in particular,given the ongoing two-way struggle), and the Pearson rank product-moment correlation coefficient (
2
) is
Article Views
Full Text
Article Tools
Related ContentIn
Science 
Magazine
Similar Articles In:
Search GoogleScholar for:
Search PubMed for:
Find Citing Articlesin:
Science Magazine
 
Science Magazine
 
Enter Search Term
 
Pattern in Escalations in Insurgent and Terrorist Activityhttp://www.sciencemag.org/content/333/6038/81.full[01.07.2011 18:45:50]
within an acceptable range for social systems (
 – 
). Although alternative progress curve forms are possible,any such two-parameter progress curve amounts to a nonlinear transformation of the power-law form andhence generates a more complex version of
.An exponential form does not generate systematicallybetter progress-curve fits;
2
for the power-law form is better by up to 70% for three-quarters of theprovinces, including (most importantly) those with the most data points, and only tends to be comparable forthe few provinces having sparse datapoints and larger scatter. We do not consider events with no deathsbecause they occur almost daily; moreover, the injury statistics are not publicly available. Although the dataresolution time scale is 1 day, this is not problematic, because the values dictating the best-fit progresscurve for a given province, or terrorist group, are usually much larger than 1.
Fig. 1
(
A
) Schematic timeline of successive fatal days shownas vertical bars. is the time interval between the firsttwo fatal days, labeled 0 and 1. (
B
) Successive timeintervals between days with IED fatalities in theAfghanistan province of Kandahar (squares). On thislog-log plot, the best-fit power-law progress curve is bydefinition a straight (blue) line with slope ( is anescalation rate). (
C
) The solid blue line shows bestlinear fit through progress-curve parameter valuesand for individual Afghanistan provinces (blue squares)for all hostile fatalities (all coalition military fatalitiesattributed to insurgent activity). The green dashed line shows value , which is the situationin which there are no correlations. The subset of fatalities recorded in icasualties as “southernAfghanistan” is shown as a separate region because of their likely connection to operations nearthe Pakistan border.
 reveals a surprising linear relationship between the best-fit progress curve values and forindividual provinces. The straight line through the provinces has
2
= 0.9 and is given by the equation, with best-fit parameter values
= 0.89 and
= −1.22. Even if one speculated that the
and values have to lie somewhere in this range (which, a priori, they do not), a simple Monte Carlocalculation puts the probability of this linear relationship emerging by chance with a comparable
2
at lessthan 1 in 10
6
.
 shows that the linear relationship extends to a specific weapon [that is, fatalitiescaused by improvised explosive devices (IEDs)] and to a completely different insurgent conflict (Iraq), as wellas to global terrorism and suicide bombings by individual groups. Theseand
values suggest, for example,that common organizational processes underlie all hostile military fatalities in Farah and Kunar, IED militaryfatalities in Kabul, and fatal terrorist attacks. The fact that
shows larger scatter than
isunderstandable given its inclusion of events in separate countries and the sparser number of IED fatalities ascompared to total hostile fatalities. Although the Afghanistan provinces exhibit an escalation of fatal daysthroughout the entire data set (that is, decreasing), the Iraq provinces cease to show pure escalationtoward the end of the conflict, where fatalities cease. Because this paper focuses on escalation, the Iraq datapoints in
correspond to the initial escalation up to the turning point in the cumulative moving average of. Other reasonable methods for determining a turning point retain the same linear structure as that in
.
Fig. 2
The solid black line shows the best linear fit throughprogress-curve parameter values and for IEDfatalities in provinces in Afghanistan (blue squares) andIraq (red squares). Because the frequency of IEDfatalities for provinces sitting just below is low,their values are relatively uncertain and should beregarded simply as . The green dashed line is at. These findings are consistent with results forIED data aggregated over all Afghanistan (blue star) andfor global terrorism and suicide bombings. Global terrorism: The dark diamond shows the andvalues deduced from the best-fit progress curve for global terrorist group activity when averagedover all attacks (
); the light diamond is an alternative estimate where and are calculateddirectly by inserting the time intervals between initial attacks into the progress curve formula(SOM). Suicide bombings: The blue triangle is for Hezbollah suicide attacks, and the white triangleis for suicide attacks within Pakistan (data are fromhttp://cpost.uchicago.edu/ ).We considered the possibility that these results derive from contagion or diffusion effects due to proximity. Thedata points for Kandahar, Kunar, and Farah sit nearby in
,showing that their escalation patterns arealmost identical; however, they are widely separated geographically within Afghanistan, with Kandahar in thesouth, Kunar in the east, and Farah in the west. If fatal days in different provinces were synchronized incalendar time (which they are not), they would sit at the same point in
,but they do not. Likewise, if
fatal days in different provinces were independent, then the individual (τ
1
,
) points should be scatteredrandomly across
,but they are not. Comparing Fig. 1C to detailed sociotechnical maps (
), we couldnot detect a clear systematic relationship between the data point locations and effects such as proximity toPakistan, density of internally displaced persons, common tribal warlords, or levels of poppy production (
). In
My Science
More InformationMore in Collections
Related Jobs fromScienceCareers
 
Pattern in Escalations in Insurgent and Terrorist Activityhttp://www.sciencemag.org/content/333/6038/81.full[01.07.2011 18:45:50]
terms of increases in troop or insurgent numbers, it would have required a very complex and specificprovince-dependent increase, combined with a very contrived mathematical relationship to fatalities, toproduce the observed patterns. Instead, the calendar times for the
’th fatal day vary wildly between provinces(for instance, the first recorded military fatality in Paktia was 4 March 2002, whereas in Wardak it was 25 July2007). Nor are the observed patterns simply linked to an increase in the number of troops and hence to anincrease in the number of targets, because monthly troop increases in Afghanistan were almost linear, not apower curve (
,
). Furthermore, as observed in Iraq after the surge, significant troop increases can actuallydecrease the military fatality rate. We also analyzed timelines from detailed online reports concerning politicalevents (SOM); however, these timelines represented a poor fit to a progress curve (
2
= 0.2) and produced
1
,
) values (τ
1
= 12.4 and
= –0.75), which were unlike those shown in
or
2
.Our broad-brush theory does not require knowledge of specific adaptation or counteradaptation mechanisms,and hence bypasses issues such as changes in insurgent membership, technology, learning, or skill set, aswell as a need to know the hearts and minds of local residents. We regard the escalation of hostilities asrepresenting adaptation and counteradaptation in a way that is analogous to the Red Queen hypothesis ofevolutionary biology (
,
). The traditional story features the Red Queen (which in our context is aninsurgency, or terrorist group) running as fast as she can just to stay at the same place, implying that her BlueKing opponent (the coalition military or antiterrorism organization) instantaneously and perfectly counteradaptsto her advances so that they are always neck and neck (the distance between the two
= 0). However,instantaneous and perfect counteradaptation is unrealistic; indeed, complex adaptation-counteradaptationdynamics generated by sporadic changes in circumstances imply that
’s temporal evolution is likely to be socomplex that it can be modeled as a stochastic process (
,
). We do not need to know exactly whychanges at any specific moment, nor do the changes in have to have the same value, because eachchange is the net result of a mix of factors [such as learning by experience or changes in personnel andtechnology (
 – 
)] for each opponent. When the Red Queen is in the lead (
> 0), there will inevitably be fataldays. If her lead increases, the rate of fatal days should also increase, hence for
> 0, we take theinstantaneous rate of fatal days as proportional to . As tends to 0, fatal days become increasingly rare.When
< 0
,
the Blue King holds the lead and there are no fatal days. By definition, provinces with no fataldays do not appear in
and
2
 (mathematically, becomes infinite and is undefined). The provincesin
and
2
exhibit escalation and are therefore consistent with an
that is generally positive andincreasing over time. For these provinces, significant changes in the Red Queen’s lead are likely to occuraround fatal days, making a function of : The Red Queen has by definition become successful at thatmoment and this may stimulate a further increase in , whereas the Blue King’s failure may stimulate aneffective counteradaptation effort and hence reduce . In the simple limiting case where changes in arestatistically independent and have finite variance, the Central Limit Theorem (
) says that the typicalmagnitude (the root mean square) of the Red Queen’s lead after steps varies as , where (SOM),which is the random walk result widely used in physics, biology, and economics (
 – 
). For the moregeneral and realistic case in which changes in depend on prior history, research on correlated walks showsthat will deviate from but will typically remain in the range (
 – 
), in agreement with
and
2
. Because the rate is inversely proportional to the time interval, this gives τ
=
an 
 –b
, where
is aconstant of proportionality. Setting
yields τ
1
=
, hence our theory reproduces the empirical finding thatthe escalation of fatal days follows an approximate power-law progress curve with typical values inthe range . Our theory interprets provinces with values near 1 as having an ineffective militarycounteradaptation, so that persistently increases at every step , hence
 
with
.
This is analogousto the Red Queen moving steadily forward at constant velocity while the Blue King remains stuck at thestarting line. If the Red Queen gains momentum,
can start accelerating and hence as observed forsome provinces such as Paktia and Ghazni in
.The fact that values for IEDs appear to beconcentrated between 0 and 1 (
) whereas the all-hostile range is from 0 to 1.5 (
) suggests thatthe coalition military counteradapts to IEDs better than to other threats. Provinces at (Kabul and Zabul in
) have a constant rate of fatal days, implying that the military is managing to contain furtherescalation
.
Similar interpretations can be made for any other data point in
and
2
,according to itsvalue.
The linear relationship between τ
1
and
in
and
2
can be reproduced by adding a coupling betweenthe Red Queen–Blue King arms races in different provinces. This coupling can arise for various reasons: Eventhough the insurgent groups (Red Queens) operating in two different locations may be unconnected, and eventhough they inflict military fatalities at different points in space and calendar time, the Blue King opponent isessentially the same for each (the same military coalition), and hence the insurgents in one province mayeventually adopt successful tactics from another. On the Blue King side, the widespread use of similar tacticsor equipment could generate coupling. For example when MRAPs (Mine Resistant Ambush Protectedvehicles) were moved to Afghanistan, the Blue King effectively took on very similar operational characteristicsin each country: slow, lumbering movement restricted to certain transport corridors.To illustrate the potential of our findings, suppose that a quiet region suddenly experiences 2 fatal days. Wecan estimate the escalation ratefor future fatal days in that region by taking the initial time interval to beand reading off the corresponding value from
 or
2
. An estimate for the date of the ’th future fatalday can also be obtained by summing all time intervals using .
 suggests that
actual time intervals may exhibit significant scatter around the τ
estimates; however, their sum (and hence
the actual calendar date) will show less variability because positive and negative errors in τ
tend to cancelwhen summed. Inputting only the first data point in
, the error between the predicted and actual timeof the ultimate fatal day (
= 92, which occurred on 13 May 2010) is less than 25%, which is far smaller than
the scatter in individual τ
values might have suggested. Access to detailed classified data sets would enablefuture analysis to fully explore this possibility, and may help establish a systematic approach to quantifying an

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->