RAND Corporation. Social Science for Counterterrorism
Подождите немного. Документ загружается.
Social-Science Foundations for Strategic Communications 325
explicitly lay out which communications strategies tend to work and
which do not. is paper aims to extract such lessons from social-science
research, supplementing the primarily observational lessons from ter-
rorism and insurgency communications with empirical research from
analogous policy efforts to prevent or deter high-risk or illegal behaviors
(for example, in health policy, education, and criminology) and lessons
from the worlds of public relations, government-sponsored broadcast-
ing, and political communications. Of course, the findings of such
studies are not statistically generalizable to the problems of terrorism or
insurgency. Nevertheless, they are useful in suggesting generally prom-
ising approaches, noting specific cross-cultural mistakes, warning of
other potential pitfalls, and highlighting research questions for future
examination.
It can be difficult to talk about “what works” in strategic commu-
nications because the practice is often confused with careless or heavy-
handed propaganda. A more subtle problem is that strategic commu-
nications operates on hugely varying time scales; what might move
the needle in daily public opinion polls might not have any effect on
long-term strategic relationships, and what might be a useful long-term
strategy, such as education, may be very hard to justify using short-
term metrics.
As a result, the remainder of this paper divides the discussion
into three sections, corresponding with the three time frames of strate-
gic communications identified by the DSB: “short-term news streams,”
“medium-range campaigns on high-value policies,” and “long-term
engagement” through relationships and dialogue (DSB, 2008). Gen-
erally speaking, the first section explores lessons in media relations
and crisis communications; the second reviews lessons for designing,
launching, and evaluating a communications campaign; and the third
reviews literature on relationship-building and long-term changes in
community norms. e discussion is not meant to exhaustively cover
all of the relevant social science but rather to highlight potentially
useful theory and practice for CT/COIN efforts.
326 Social Science for Counterterrorism: Putting the Pieces Together
Short-Term Communications
In countering terrorism or insurgency, short-term communications
efforts—such as crisis response or rapid message adaptation in the
face of changing events—can affect many of the key factors identi-
fied in companion papers. For instance, the communications response
to news of collateral deaths during a security operation might influ-
ence perceived grievances among the general population, whereas the
response to a particularly violent terrorist attack might affect the per-
ceived legitimacy of terrorism. Most of the prescriptive literature in this
area focuses on the former type of events, that is, crises or news events
that pose a threat to the reputation or legitimacy of those doing the
communicating.
Planning for Short-Term Crises
One common prescription for short-term communications is that
organizations should plan for crises and other surprises ahead of time.
Within the literature on crisis communications is a generally agreed-
on list of preparatory actions an organization should take, including
writing a crisis response plan, assigning crisis spokespersons and teams,
training staff for crises, identifying communications risks and vulner-
abilities, maintaining media contact lists, and monitoring the media
for key events (Borda and Mackey-Kallis, 2003; Cloudman and Hal-
lahan, 2006; Gainey, 2006). ere has been some empirical research
on which organizational factors are associated with successful comple-
tion of the above tasks; among these factors are an autonomous com-
munications department, an environment encouraging the delegation
of authority, a philosophy that is process-oriented (as opposed to, for
instance, outcome-oriented), and past experience with crises (Cloud-
man and Hallahan, 2006; Guth, 1995).
Study of whether or not these tasks actually predict crisis com-
munications “success” has been primarily observational. Although the
bulk of reviews conclude that planning is effective, others find that
crisis plans are weaker predictors of crisis outcomes than simply having
a proactive, aggressive, “tell our side of the story” organizational cul-
ture (Marra, 1998) or following a consistent code of values (Fitzpatrick,
Social-Science Foundations for Strategic Communications 327
1995). In an attempt to distinguish why some plans work and some do
not, one reviewer (Fearn-Banks, 2007) argued that communications
planning is ineffective when there has been a failure of imagination
(the particular type of crisis was not anticipated) or a failure to regu-
larly update the plan.
Methods of Response
Within the communications literature, some attempt at short-term
response to crises and rumors—particularly those that threaten an
organization’s reputation or legitimacy—is, not surprisingly, consid-
ered essential (Borda and Mackey-Kallis, 2003; Gainey, 2006; May-
nard, 1993). is observational lesson has theoretical roots in the fun-
damental attribution error (Jones and Nisbett, 1972), which states that
when an observer watches an actor commit a negative behavior—but
does not know the full context behind the behavior—the observer
will too easily attribute the behavior to negative traits of the actor.
In other words, a civilian who notices longer wait times at a security
checkpoint—but is not aware that this is in response to a recent foiled
attack—is likely to attribute the change to simple malice or insensitiv-
ity among the security forces. erefore, one role of short-term com-
munications is to provide the public with the contextual information
behind a crisis response so that the response is not simply attributed to
ill will. Two other roles, detailed below, are in responding to fact-based
crises (that is, crises rooted in a mistake, controversial action, or unex-
pected event) and in responding to controversies based in rumor and
disinformation.
Response to Fact-Based Crises. In the terrorism and insurgency
contexts, one example of this crisis might be the public outcry after
a security operation results in collateral civilian injuries. Public rela-
tions literature has addressed this topic with such theories as situational
crisis communication theory (Coombs and Holladay, 2002) and the
communicative response model (Bradford and Garrett, 1995), which
match different response types to different crisis types. For instance,
the latter model lays out four crisis responses and matches each with
appropriate crisis conditions: denial (when actors can show they did
not commit the offense), excuses (when actors clearly committed the
328 Social Science for Counterterrorism: Putting the Pieces Together
offense but can show they were not in control of the situation), justi-
fication (when actors were clearly in control of the situation but there
were no clear standards of right and wrong), and concession (when
clear standards were violated).
Empirical lessons from an initial quantitative study of this frame-
work were that failure to give any response at all had a clear negative
effect on the organization’s image, and, contrary to theory, conces-
sion was an optimal response for maintaining one’s image, even where
theory implied that concession was not appropriate (for example, when
there was evidence that could have been used to make a denial) (Brad-
ford and Garrett, 1995). In dealing with the culturally sensitive crises of
terrorism and insurgency, the power of concession and apology might
be even greater. However, later research in this same vein found that
using the situation-appropriate crisis response as predicted by theory
was correlated with positive subsequent media coverage of the crisis
(Huang, 2006).
Taking a slightly different perspective, other studies (Coombs,
1999; Coombs and Holladay, 2006) found that an offending organiza-
tion’s offer of sympathy and compassion (as opposed to instructional
messages, formal apologies, or victim compensation) led to a more sym-
pathetic and less-angry audience and that apologies delivered by print
media were received slightly better than identically worded apologies
delivered by video. Truth and completeness of the communications
response, of course, is also strongly recommended in the literature; sur-
veys have found that 95 percent of individuals are more angered by
lying about a crisis than they are by the actions precipitating the crisis
itself (Maynard, 1993). Audience reaction to these messages can also
be affected by such factors as spokesperson attractiveness and ethnic-
ity; the more that audiences judge a crisis spokesperson to be similar to
themselves, the more likely they are to perceive the crisis communica-
tion as credible (Arpan, 2002).
A final dimension to crisis response is the speed with which it
is delivered. A recent review of crisis research concluded that rapid
response is a necessary but insufficient condition for successful crisis
communications (Borda and Mackey-Kallis, 2003). Although some
studies have quantified the required response speed—concluding, for
Social-Science Foundations for Strategic Communications 329
instance, that an organization has up to 12 hours after a story breaks
to gain control of the message (Gainey, 2006; Small, 1991)—most lit-
erature discusses the benefits of speed in relative terms, that is, get-
ting the organization’s “story” out faster than competing narratives.
However, greater response speed may have drawbacks. A recent experi-
ment involving journalists (Arpan and Pompper, 2003) found that if
an organization rushes to break a story to the media before the media
discovers the story from a third party, the media will rate the organiza-
tion as more credible, but the media will also be more interested in the
story than it would have been otherwise.
Response to Rumors and Disinformation. is type of response
has always been a necessary part of wartime strategic communications;
for instance, the failure to rebut rumors of U.S. germ warfare during the
Korean War (in the mistaken belief that the rumors would disappear
on their own) has been considered a clear mistake (Shaw, 1999). Coali-
tion forces in Iraq were the subject of potentially damaging rumors (for
example, that U.S. soldiers distribute pornography to children) as early
as the spring of 2003 (Hendon and Holton, 2003).
e literature on rumor psychology, and more specifically rumor
management, offers a number of useful insights. A meta-analysis of
prior theory and empirical research (Rosnow, 1991) identified three
necessary factors for the dissemination of rumor: general uncertainty
about the rumored issue, high personal anxiety levels, and the believ-
ability of the particular rumor. Consequently, rumors and enemy dis-
information can be attacked by weakening any of these three factors
(for example, offering greater information to the public to quell uncer-
tainty, or attacking the credibility of a rumor source). Not surprisingly,
spokespersons perceived as honest are most effective in reducing rumor
anxiety and believability; although a more knowledgeable or high-
status spokesperson will further enhance this effectiveness, knowledge
and status alone are insignificant in the absence of perceived honesty
(Bordia, DiFonzo and Schulz, 2000).
Other survey-based research (DiFonzo and Bordia, 2000) con-
cluded, among other findings, that two strategies—rumor-control
hotlines and denial of rumors by trusted third parties—are generally
underused relative to their perceived effectiveness. Vigorous refutations
330 Social Science for Counterterrorism: Putting the Pieces Together
of rumors may be most effective when the source of the rumor or dis-
information is portrayed as having something to gain from the rumor
and when those doing the refuting end their message on a conciliatory
note (Iyer and Debevec, 1991). As for how best to transmit these refu-
tations, U.S. sources on the ground in Iraq have argued that face-to-
face rumor control is much more effective than print or radio efforts
(Steele, 2003). However, close monitoring and message adaptation is
essential to ensure that the refutation does not backfire; recent political-
science research has suggested that, among committed ideologues, the
attempted refutation of a rumor only strengthens the underlying mis-
perceptions (Nyhan and Reifler, 2008).
Mid-Range Communications Campaigns
Mid-range communications consist of preplanned information cam-
paigns designed to change public attitudes in support of CT/COIN
policies (that is, attitudes toward cooperating with security forces).
Campaigns can also change behaviors: Recent research on the U.S.
public has found that the more that individuals consider terrorism to
be a serious social problem in which they are personally involved, the
more they will seek out information about it, which in turn predicts
a greater likelihood of performing recommended protective behaviors
(Lee and Rodriguez, 2008). In general, social-science contributions for
communications campaigns can be divided into three areas: formative
evaluation (that is, what to ask before launching a campaign), message
content and delivery, and summative evaluation (that is, what to ask
after the campaign ends to measure effects and outcomes).
Formative Evaluation
In the social sciences, formative evaluation has been defined simply as
“the collection of information that helps to shape the campaign” (Coff-
man, 2002). e commonsense proposition that this process benefits
subsequent program success has been empirically validated (Brown
and Kiernan, 2001), but detailed lessons on precisely how best to run
Social-Science Foundations for Strategic Communications 331
this process are, again, primarily observational in nature. e key con-
tribution of social science in this area has been in providing a list of
common questions a communications officer may wish to answer before
launching a campaign. A nonexhaustive selection of useful questions
is listed below, as drawn from a variety of evaluation literature (Atkin
and Freimuth, 2001; Barthe, 2006; Coffman, 2002; Valente, 2001;
Wixon, 1998). Evidence supporting the value of particular questions is
included where applicable.
Defining a Campaign “Theory of Action.” One formative evalu-
ation lesson confirmed in practice has been the key role of a theoreti-
cal underpinning in predicting campaign success (O’Keefe and Reid,
1990). is goes beyond borrowing from existing theory and includes
the creation of a program-specific “theory of action,” explicitly laying
out each hypothesized logical relationship linking program design to
eventual program effects (for example, a particular message will increase
awareness of issue X, which will reduce behavior Y among the public,
which will reduce the ability of terrorist groups to do Z).
One useful theoretical backbone for such an exercise is informa-
tion processing theory (McGuire, 1978), which posits a pathway of
message effectiveness through (1) audience exposure (to the message),
(2) audience attention, (3) comprehension of the message, (4) agree-
ment with the message, (5) retention, and (6) behavior change. Because
every step is necessary for success, this type of layout is useful in pin-
pointing precisely where an ineffective campaign has broken down
(Coffman, 2002; Wixon, 1998). It also highlights trade-offs in com-
munications strategy; for instance, a message may be broadly compre-
hensible at the expense of persuasiveness.
Basic questions to ask at this stage include
What is the specific knowledge, attitude, or behavior that the •
campaign is trying to change?
What is the logical or theoretical relationship between the com-•
munications effort and the desired change?
What assumptions or intermediate steps are required for the cam-•
paign to perform as predicted?
332 Social Science for Counterterrorism: Putting the Pieces Together
What are the alternative messages, themes, tactics, and strategies •
available to the campaign?
How long will the campaign last?•
Defining and Segmenting the Target Audience. Communica-
tions research generally confirms the benefits of audience segmentation.
Audiences prefer messages that are personally relevant to them and will
judge broadly targeted messages as relatively ineffective (Fishbein et al.,
2002). Conversely, national communications campaigns are improved
when they include local tags or references (O’Keefe and Reid, 1990). In
general, although segmenting by simple demographics or geography is
easiest, a campaign may be most effective when segmenting along those
factors (for example, attitudes toward suicide bombing) that underlie
the behavior to be changed (Slater, 1995). Failure to segment can be
dangerous: Different audiences can differ in their basic interpretation
of the same message (Barthe, 2006), and a particular message can be
well-received by one segment but judged as not credible by a different
segment (Skinner and Slater, 1993). However, such intergroup differ-
ences are not always as large as might be expected (Borzekowski and
Poussaint, 2000).
General segmentation questions to ask include
Will the campaign target segments of the general public, current •
or potential terrorists/insurgents, their current or potential sup-
porters, their social companions, or others?
What factors underlie the behavior to be changed? Can the audi-•
ence be segmented along those factors? If not, what alternative
segmentation strategy will be used?
Understanding Audience Segments. Segmentation is useful
only when the individual audience segments are studied and under-
stood. For campaigns that aim to affect behavioral decisions (such as
whether or not to join a terrorist group or report suspicious activity
to the police), stages-of-change models (DiClemente and Prochaska,
1985) provide a useful theoretical structure. e theory categorizes
audiences into five stages depending on their relationship with key
Social-Science Foundations for Strategic Communications 333
behavior at issue: precontemplation (the behavior is not personally rel-
evant to them), contemplation (the behavior is relevant, but they are
not prepared to perform it), preparation (they are prepared to perform
it, but have not actually done so), action (they have performed it, but
not maintained the behavior over time), or, ultimately, maintenance
of the behavior. is theory highlights the importance of fit between
message and stage (that is, individuals thinking about engaging in ter-
rorism or insurgency will likely require a very different message from
those already practicing it).
Another useful way to strengthen campaign design through audi-
ence research is in isolating the target audience’s “hot button” and
“cold button” values. is method, which has already been studied
in the context of terrorist deterrence (MacNulty, 2008), entails profil-
ing the audience to determine which hot button values are important
to them (such as material wealth or religion) and which cold button
values are not (for example, loyalty or freedom of thought). Communi-
cations campaigns can then focus on the right levers to influence atti-
tudes or behavior. is approach can also help avoid counterproductive
messaging. For example, consider differences in the degree to which
individuals value the ability to think for themselves: An information
campaign highlighting the inconsistency of joining a terrorist group
and being able to think for oneself may be effective for people who hold
this as an important value, ineffective for people who do not care about
this value, and counterproductive for people who, for whatever reason,
prefer to simply follow the decisions of others.
Audience research to consider at this stage includes
What are each audience segment’s values, beliefs, and attitudes •
toward terrorism/insurgency and CT/COIN policy?
How important are different values and desires in driving a target •
group’s terrorism- or insurgency-related behaviors?
How salient a problem is violence to each segment?•
What are the skills and self-efficacy beliefs of each segment regard-•
ing terrorism- or insurgency-related behaviors (such as joining a
group or reporting suspicious activity)?
Who are the most credible messengers for each segment?•
334 Social Science for Counterterrorism: Putting the Pieces Together
Understanding the Program Environment. Formative research
can also include analysis of on-the-ground factors that may help or
hurt the campaign, such as cultural or language barriers, relationships
with community leaders, political constraints, and staff skill levels. To
gauge how these factors might change over the lifetime of the cam-
paign, a number of increasingly sophisticated computer simulation and
prediction tools, such as the Senturion program (Abdollahian et al.,
2006), allow for the forecast of political dynamics over a mid-range
time frame (up to two years). Such programs can also strengthen for-
mative evaluation by simulating the future effects of different versions
of the communications campaign.
In general, researchers may wish to study the following items:
What resources (for example, information, staff skills, local com-•
munity support) are currently adequate and which are still needed
or otherwise unsatisfactory?
What factors in the target community will be obstacles to pro-•
gram success, and what factors will be helpful?
What do local stakeholders think about the campaign, and how •
will they react once the campaign begins?
Understanding the Media Environment. e issue of media envi-
ronment goes beyond merely determining which channels the audi-
ence is watching. However, these mundane issues should not be over-
looked; for instance, the BBC’s anticommunist broadcasts to Albania
in the early Cold War were played 90 minutes before the electricity was
switched on in Albanian towns (Defty, 2002).
Media theories, such as priming theory (Iyengar and Kinder,
1987) and the agenda-setting model (McCombs and Shaw, 1973), point
to the fundamental role of the media in determining which issues the
public deems to be important (issue salience). Increased issue salience,
in turn, leads to increased public knowledge about the issue, stron-
ger opinions on it, and a greater likelihood of participating in politi-
cal actions related to the issue (Weaver, 1991). Not surprisingly, then,
public service announcements are generally more effective when deal-
ing with topics already covered extensively by the news media (O’Keefe