A Look at Terrorist Behavior: How They Prepare,
Where They Strike
by Brent Smith, Ph.D.

Tmothy McVegh, the Sept. 11 hjack- Although we know a great deal about the
ers and Erc Rudolph all had somethng behavor of tradtonal crmnals, lttle nfor-
n common — they selected targets
maton has been avalable about terrorsts.
hundreds of mles from where they lved.
Are they much dfferent from conventonal
McVegh wandered the Mdwest lvng
crmnals, who tend to commt ther crmes
as a transent before makng hs bomb n
close to home?1 Research has shown that
Herngton, Kan., and drvng 250 mles south tradtonal crmnals are spontaneous, but
to blast the Alfred P. Murrah Federal Buldng terrorsts seem to go to great lengths prepar-
n Oklahoma Cty. The Sept. 11 hjackers
ng for ther attacks — and may commt other
traveled hundreds of mles to ther targets.
crmes whle dong so. How long does ths
And Rudolph drove nearly 300 mles from
plannng take? And do dfferent types of
Murphy, N.C., to bomb an aborton clnc n
terrorst groups vary n preparaton tme?
Brmngham, Ala.
To help answer these questons, the
For local polce departments searchng for
Natonal Insttute of Justce (NIJ) launched
ways to stop terrorst acts before they occur, a seres of projects to explore patterns
ths does not brng much comfort. When
of terrorst behavor.2 In the frst of these
lookng at these attacks, offcers mght get
projects, a panel of experts3 was assembled
the mpresson that there s not much they
to examne 60 case studes nvolvng
can do about terrorsm other than mprovng
terrorst ncdents n the U.S. durng
physcal securty at hgh-rsk targets.
the past 25 years.4 These cases nvolved
the four major types of U.S. terrorst
But were these nfamous terrorsts typcal?
groups: left wng, rght wng, sngle
2

---

N I J J o u r N a l / I s s u e N o . 2 6 0
ssue and nternatonal.5,6,7 The panel —
ncludng ths author — looked at the
Distance From Terrorist Residence to Target
homes of the terrorsts, the locatons of
(All Groups)*
plannng and preparaton, and the stes
44% (0-30 miles)
7% (31-90 miles)
8% (91-270 miles)
12% (271-810 miles)
29% (811-2,570 miles)
of the terrorst ncdents to dscover
whether any patterns emerged.
N
What we learned was ntrgung: The cases
of McVegh, the Sept. 11 hjackers and
Rudolph are actual y unusual. In fact, we
found that most terrorsts lve close to ther
selected targets, and they engage n a great
deal of preparaton — some over the course
of months or even years — that has the
potental of comng to the attenton of local
law enforcement.
Terrorists Think Globally
W
E
but Act Locally



We studed:









■ Ten attacks by nternatonal groups that








30 m
nvolved 93 preparatory acts.

i

les










9

0

miles
■ Fourteen attacks by rght-wng groups that


nvolved 55 preparatory acts.





2

7

0

m

i

les


■ Twenty-nne attacks by envronmental
810 miles
groups that nvolved 80 preparatory acts.
2,430 miles
■ Sx attacks by left-wng groups that
nvolved eght preparatory acts.
S
* Ths s a lnear analyss of the dstance from the resdences of 423 terrorsts to
Accordng to our analyss, almost half
ther targets.
(44 percent) of all terrorsts examned lved
wthn 30 mles of ther targets. (See spa-
targets. Terrorsts may stay close to home
tal analyss on ths page, “Dstance From
because of new mmgraton status, lack
Terrorst Resdence to Target (All Groups).")
of transportaton, lack of knowledge of
When the types of terrorst groups are
the urban landscape or a desre to avod
examned separately, however, the fnd-
attenton. Among sngle-ssue terrorsts
ngs are much dfferent.
n partcular, 71 percent of the preparatory
acts occurred wthn 12 mles and 92 percent
Internatonal terrorsts lved relatvely near
wthn 28 mles of the target. Ths fndng may
ther targets, whereas rght-wng terrorsts
also be attrbuted to the use of “uncoord-
lved n rural areas but selected targets
nated volence” tactcs by these envronmen-
reflectng the “pollutants of urban lfe”
tal and ant-aborton extremsts, whch often
n nearby ctes.
results n local targetng by “lone wolves”
sympathetc to the cause.
Terrorsts most commonly prepared for ther
attacks wth survel ance and ntel gence
A separate fol ow-up NIJ project8 that
gatherng, robberes and thefts to rase fund-
analyzed the dstance between more
ng for the group, weapons volatons, and
than 250 envronmental and nternatonal
bomb manufacturng. Agan, most of these
terrorsts’ homes and ther targets confrmed
behavors took place relatvely near ther
the earler prelmnary fndngs that ther
homes, whch, n turn, were close to the
spatal patterns are farly smlar. The analyss
3

---

N I J J o u r N a l / I s s u e N o . 2 6 0
of the attack. Major crmes to procure fund-
Distance From Environmental Terrorist
ng for the group — lke thefts, robberes
Residence to Target*
and burglares — however, are ntentonally
commtted many mles away to avod
51% (0-30 miles)
9% (31-90 miles)
8% (91-270 miles)
8% (271-810 miles)
24% (811-2,430 miles)
drawng attenton to the group’s locaton
N
and target choce.
The Terrorist’s Timepiece
We found that preparatons general y began
less than sx months before the attack and
ended wth a flurry of actons a day or so
before. Ths pattern vared by group type.
Sngle-ssue and rght-wng terrorsts
engaged n substantal y less preparatory
crme over a shorter perod — once agan,

most lkely reflectng the use of “leader-


W
E
less resstance” and lone-wolf strateges.

The plannng cycle of nternatonal terrorsts

tended to be longer. (See the table on page




5, “Cumulatve Percentage of Preparatory










Acts Over Tme.”)


3

0

mi


le


s









9

0



m
In our fol ow-up study, we took a closer

ile


s



2
look at the specfc patterns of nternatonal


7



0




m




i


le



s
and envronmental terrorsts by placng





8

1

0



m

the preparatons for al ncdents on a

i

l
2
es
,430
tme lne. For nstance, we examned the
miles
21 ncdents attrbuted to the envronmen-
tal terrorst group known as “The Famly,”
S
whch was responsble for the Val, Colo.,
sk resort arson n 1998 and many attacks
* Ths s a lnear analyss of the dstance from the resdences of 208 envronmental
terrorsts to ther targets.
aganst Forest Servce and Bureau of Land
Management buldngs from 1996 to 2000.
found that about half of the envronmental
The Famly conssted of at least 16 people.
terrorsts and nearly three-ffths of the nter-
Unlke most envronmental terrorsts who use
natonal terrorsts lved wthn 30 mles of
uncoordnated volence and lone-wolf strate-
ther targets. (See the fgure on ths page,
ges, the group’s actons were more sponta-
“Dstance From Envronmental Terrorst
neous than other envronmental cases, wth
Resdence to Target” and the one on page
a short preparaton perod and lttle extended
5, “Dstance From Internatonal Terrorst
plannng. Eghty-fve percent of ther known
Resdence to Target.”) Sxty-fve percent of
preparaton actvtes — typcal y, nspecton
the envronmental terrorsts and 59 percent of of the target, purchase of bomb-makng
the nternatonal terrorsts prepared for ther
tems from local stores and dentfcaton
attacks wthn 30 mles of ther target stes.
of a stagng area a short dstance from the
target — occurred wthn sx days of the
Although the terrorsts studed commtted
planned attack. An explosve devce was
most of ther preparatory offenses near ther assembled at the stagng area a day or so
homes, they conducted robberes, burglares before the ncdent and then delvered to the
and thefts much farther away — an average
target. Partcpants usual y returned to the
of 429 mles from home. Ths suggests that
stagng area to destroy any evdence.
most envronmental and nternatonal terror-
sts lve near the selected target and conduct Internatonal terrorsts, on the other hand,
survellance and other general preparaton
engaged n nearly three tmes as many
near ther homes and the eventual locaton
preparatory acts per ncdent as ther
4

---

N I J J o u r N a l / I s s u e N o . 2 6 0
envronmental counterparts. Ths may be
due to the larger number of people usually
Distance From International Terrorist
nvolved n nternatonal ncdents, the sze
Residence to Target*
and scope of the planned ncdent or smply
59% (0-30 miles)
17% (31-90 miles)
8% (91-270 miles)
12% (271-810 miles)
4% (811-2,430 miles)
a longer plannng cycle. Comparng the 10
nternatonal terrorst ncdents that occurred
N
on Amercan sol,9 we found that the aver-
age plannng cycle for nternatonal terrorsts
was 92 days, as opposed to 14 days for
envronmental terrorsts. Averages can be
msleadng, however, because of sgnfcant
outlers, such as the multyear plannng cycle
of the Islamc extremsts seekng to destroy
New York Cty landmarks n the md-1990s.
Whereas envronmental terrorsts com-
mtted an overwhelmng majorty of ther
preparatory actvtes n the week before the
ncdent, nternatonal terrorsts took up to
W
E
sx months to prepare.
Arming Police With Knowledge

For law enforcement agences, the mpl-
30 miles

catons of these patterns are sgnfcant.

90 miles
Commttng an act of terrorsm wl usual y
nvolve local preparatons. Although much

270 miles
of ths conduct wl not necessarly be crm-
nal, early ntel gence may gve law enforce-
810 miles

2,430 miles
ment the opportunty to stop the terrorsts


before an ncdent occurs. Knowledge of the
threat — for example, understandng how
S
long envronmental or nternatonal terror-
* Ths s a lnear analyss of the dstance from the resdences of 49 nternatonal
sts prepare for ther attacks — wl affect
terrorsts to ther targets.
the manner n whch local offcals respond.
Identfyng preparatory actons by envron-
Understandng that most terrorsts “act
mental extremsts may sgnal that an attack
local y” can be mportant to know as nves-
s mmnent, whereas smlar behavor by an
tgatve agences seek to prevent terrorsm
nternatonal group mght suggest that an
and arrest perpetrators. These local patterns
attack s stl several months away.
may be used by agences to more effcently
Cumulative Percentage of Preparatory Acts Over Time
Type of Terrorist Incident
Day
2-3
4-7
8-14
15-30
1-3
4-6
7-12
1-3
3+
Group
day
before
days
days
days
days
mos
mos
mos
yrs
yrs
International
5%
9%
13%
21%
32%
55%
68%
84%
89%
97%
100%
Right wing
15%
26%
31%
44%
49%
67%
94%
96%
96%
100%
Environmental
20%
43%
72%
80%
91%
95%
98%
100%
Left wing
50%
75%
88%
88%
88%
88%
100%
All categories
13%
27%
40%
48%
57%
72%
85%
93%
95%
99%
100%
Percentage of acts completed wthn a specfed tme range:
0-25%
26%-50%
51%-76%
77%-100%
5

---

N I J J o u r N a l / I s s u e N o . 2 6 0
patrol known, hgh-rsk target areas and
gather ntellgence on suspected actons
AcKNoWLEDGmENTs
wthn a specfc dstance from potental
We would lke to express our apprecaton to
targets. As we contnue to deepen our
the FBI, the U.S. House of Representatves
understandng of the relatonshp among
Judcary Subcommttee on Crme, the U.S.
the locaton of the terrorst’s home, terror-
Senate Judcary Commttee, Sen. Jeff
st preparaton actvtes and the target, ths
Sessons, the Natonal Memoral Insttute for
growng knowledge should help offcers
the Preventon of Terrorsm n Oklahoma Cty,
prevent and respond to attacks.
the Department of Homeland Securty and NIJ
NCJ 222900
for supportng our research and the Amercan
terrorsm study.
For More Information
terrorsm; Steve Chermak, rght-wng terror-
■ More nformaton on ths study s avalable
at http://trc.uark.edu.
sm; Kelly Damphousse, terrorsm database
management; Wllam Dyson, domestc and
■ For more nformaton on NIJ’s terrorsm
nternatonal terrorsm; Mark Hamm, rght-
research, see http://www.ojp.usdoj.gov/
wng terrorsm; Robert Hebel, left-wng ter-
nj/topcs/crme/terrorsm.
rorsm; Austn Turk, terrorsm and poltcal
volence theory; and the author, Brent Smth,
domestc and nternatonal terrorsm.
Notes
4. Cases were selected prmarly from the
Amercan Terrorsm Study, a project that
1. Wrght, R., and S. Decker, Burglars on the
nvolves data collecton on all persons ndcted
Job: Streetlife and Residential Break-ins,
as a result of an FBI “terrorsm enterprse”
Boston: Northeastern Unversty Press, 1994;
nvestgaton as defned by the attorney
Wrght, R., and S. Decker, Armed Robbers in
general gudelnes for domestc and nterna-
Action: Stickups and Street Culture, Boston:
tonal terrorsm nvestgatons.
Northeastern Unversty Press, 1997; and
Reppetto, T. A., Residential Crime, Cambrdge:
5. Left-wng terrorst groups generally refer to
Bal nger Publshng, 1974.
those that adhere to a “forward-lookng”
deology, one that advocates a poltcal or
2. “Pre-Incdent Indcators of Terrorst Group
socal system that has not exsted before n
Actvtes,” NIJ Grant # 2003-DT-CX-0003;
the U.S. Typcally assocated wth extreme
“Geospatal Analyss of Terrorst Actvtes:
lberalsm, examples nclude the May 19th
The Identfcaton of Spatal and Temporal
Communst Party, the Weather Underground
Patterns of Preparatory Behavor of Inter-
and the Black Lberaton Army.
natonal and Envronmental Terrorsts,” NIJ
Grant # 2005-IJ-CX-0200; and “Terrorsm
6. Rght-wng terrorst groups generally refer to
n Tme and Space,” NIJ Grant # 2006-IJ-
those that adhere to a “backward-lookng”
CX-0037.
deology, one that advocates a return to a
poltcal or socal system that s perceved to
3. The members of the panel and ther areas of
have exsted prevously n the U.S. Typcally
expertse were: Ron Arnold, envronmental
assocated wth extreme conservatsm,
examples nclude the KKK, whte supremacy
groups lke the Aryan Natons or groups lke
the Sherff’s Posse Comtatus that oppose
About the Author
centralzed federal power.
Brent Smith is a professor of sociology and criminal justice at the
7. Lke the name mples, sngle-ssue terror-
University of Arkansas. A student of terrorism for nearly 30 years,
st groups advocate on behalf of a partcular
he created the American Terrorism Study in 1988 with assistance
poltcal or socal ssue, such as ant-aborton
or the envronment.
from the FBI. Smith currently serves as director of the Terrorism
Research Center (TRC) in Fulbright College at the University of
8. “Geospatal Analyss of Terrorst Actvtes,”
Arkansas. He was assisted on the projects discussed in the article
NIJ Grant # 2005-IJ-CX-0200.
by Kelly Damphousse, professor of sociology and associate dean
9. Temporal data on nternatonal terrorsts
of the College of Arts and Sciences at the University of Oklahoma;
are lmted due to the smal number of nter-
Jackson Cothren, assistant professor of geosciences and affiliate
natonal ncdents that have taken place n the
U.S. Because of the FBI’s success n dsruptng
of the Center for Advanced Spatial Technologies at the University
plots, the number of cases s smal .
of Arkansas; and Paxton Roberts, research associate at TRC.
6

---

Document Outline

• 222899 6
• 222899 7
• 222899 8
• 222899 9
• 222899 10

---