Crowd control report

Crowd Behavior, Crowd
Control, and the Use of
Non-Lethal Weapons

HUMAN
EFFECTS
ADVISORY
PANEL
Report of Findings

Dr. John M. Kenny
Dr. Clark McPhail
Dr. Peter Waddington
Lt. Sid Heal
Maj. Steve Ijames
Dr. Donald N. Farrer
Dr. Jim Taylor
Capt. Dick Odenthal INSTITUTE FOR NON-LETHAL
1 January 2001 DEFENSE TECHNOLOGIES

Institute for Non-Lethal
Defense Technologies

Human Effects Advisory Panel
Report of Findings

Crowd Behavior, Crowd Control,
and the Use of Non-Lethal Weapons

Dr. John M. Kenny
Dr. Clark McPhail
Dr. Peter Waddington
Lt. Sid Heal
Maj. Steve Ijames
Dr. Jim Taylor
Capt. Dick Odenthal

1 January 2001

This material is based upon work supported by the United States Marine Corps under
Contract No. M67004-99-D-0037, Delivery Order No. 0001.

Any opinions, findings, and conclusions or recommendations expressed in
this material are those of the author(s) and do not necessarily reflect
the views of the United States Marine Corps and should not be construed
as an official Department of Defense position, policy, or decision.

Approved for Public Release, JNLWD, 20 Feb 2004

Contents
EXECUTIVE SUMMARY ...................................................................................... 1
What Is Known and Not Known About Crowd Behavior ....................................... 1
The Negotiation Model: A Decision-Making Guideline for Crowd Control ............. 1
Pre-incident Planning ..................................................................................... 1
Background Information and Intelligence ...................................................... 2
Encouraging Relationships and Ownership ................................................... 2
Acceptability—Redefining What it Means to Win ........................................... 2
Education and Training Guidelines for Crowd Control ........................................... 2
A “Road Map” for Future Research on Crowd Behavior ....................................... 2
Why do people assemble? ............................................................................ 2
What organizational structures exist within a gathering? .............................. 3
What determines the intensity or “energy” of a gathering? ............................ 3
How do crowds in other cultures behave? ..................................................... 3
What psychological and social factors contribute to violent
behavior in crowds? ....................................................................................... 3
What cultural and situational factors contribute to violent crowd behavior? .. 3
What military or law enforcement actions lead to violent crowd behavior? .... 3
What crowd-control techniques are most effective in preventing
and stopping violent behavior? ...................................................................... 3

INTRODUCTION .................................................................................................. 4
Purpose ................................................................................................................ 4
Requirements for Crowd Control .......................................................................... 4
The Panel .............................................................................................................. 5
Preparation for the Panel ...................................................................................... 6
The History of Hostile Crowds and Riots .............................................................. 6
Conduct of the Panel’s Assessment ................................................................... 10

OBJECTIVE 1
What Is Known And Not Known About Crowd Behavior .............................. 12
General ............................................................................................................... 12
Discussion ........................................................................................................... 12
Debunking Stereotypes ............................................................................... 12
The Crowd: A Process—Not An Entity ........................................................ 13
The Assembling Process ............................................................................. 14
The Temporary Gathering ............................................................................ 15
The Dispersing Process ............................................................................... 16
What We Know About Riots ......................................................................... 18
Frequency of Riots ....................................................................................... 18
Community Demographics .......................................................................... 19
Participants .................................................................................................. 19
Conclusion ................................................................................................... 21

OBJECTIVE 2
Decision-making Guideline for Crowd Control .............................................. 22
General ............................................................................................................... 22
Discussion ........................................................................................................... 22
Pre-incident Planning ................................................................................... 22
Background Information and Intelligence .................................................... 23
Encouraging Relationships and Ownership ................................................. 27
Acceptability—Redefining What It Means to Win ......................................... 30
Conclusion .......................................................................................................... 32

OBJECTIVE 3
Development of Education and Training Guidelines for Crowd Control ...... 34
General ............................................................................................................... 34
Discussion ........................................................................................................... 34
The Need for Case Studies ......................................................................... 34
How to Use Case Studies ............................................................................ 35
Knowledge Management – Lessons Learned ............................................. 36
Conclusion .......................................................................................................... 37

OBJECTIVE 4
Development of a “Road Map” for Future Research on Crowd Behavior .... 38
General ............................................................................................................... 38
Discussion ........................................................................................................... 38
The Need for Research ............................................................................... 38
A “Road Map” for Future Research .............................................................. 39
Conclusion .......................................................................................................... 41

CONCLUSION .................................................................................................... 42

Sidebar text highlights
salient points throughout
the report.

Executive Summary
The Human Effects Advisory Panel convened from 12-14 September 2000. Its
purpose was to assess crowd behavior and the potential for crowd control. This is
a leading core capability sought by the Joint Non-lethal Weapons Program. The
need to thoroughly examine crowd behavior grew out of the Panel’s previous
assessments of non-lethal weapons, which are being designed for crowd control.
The following is a summary of the Panel’s findings.

What Is Known and Not Known About Crowd Behavior
Military and law enforcement concepts on crowd control are based on stereotypes
that have been questioned by social scientists. These scientists have gathered
empirical data, indicating that:
s Crowds are not homogeneous entities—all participants are not the same.
s Crowds are not made up of isolated individuals, but of “companion
clusters”, which arrive, remain and leave together.
s Crowd participants are not unanimous in motives.
s Crowd participants do not necessarily assume a sense of anonymity.
s Crowds are not given to unique emotional displays.
s Crowd participants seldom act in unison and, if they do, it does not last long.
s Crowds do not cripple individual cognition.
s Crowds are more of a process—they have a beginning, middle and end.
s Crowds are not uniquely distinguished by violence.

Social, political and economic factors are not consistent predictors of riot intensity.
Military and law enforcement must rethink crowd control. New approaches should
be based on this empirical evidence as well as the practical experiences of the
law enforcement community. Additionally, this information should serve as the basis
for new training and decision-making protocols for crowd control.

The Negotiation Model: A Decision-Making Guideline
for Crowd Control
To facilitate a new approach, the Panel developed a decision-making guideline for
crowd control based on the development of a negotiation model. The goal of this
guideline is to avoid confrontation. However, the guideline does not provide answers.
Rather, it offers the precepts and key questions which practitioners and
academicians believe are basic to crowd control.

Pre-incident Planning
This planning begins well before the incident. It includes strategic planning on
crowd control as well as responsibilities, training, organization, operating

Crowd Behavior, Crowd Control

1

procedures, and rules of engagement. This planning emphasizes avoidance of
confrontation and prevention of serious outcomes in crowd control situations. Above
all, it focuses on the question: “What are we are trying to accomplish?”

Background Information and Intelligence
“What we seek to accomplish” must continually be assessed against what others
seek to accomplish. The assessment of crowds has its own set of questions, which
this guideline provides. The guideline offers suggestions for when and how to
intervene in violent individual or collective behaviors when they occur in a crowd.

Encouraging Relationships and Ownership
Working relationships between tactical commanders and group leaders are the best
means of preventing bad outcomes in crowd situations. This aspect of the guideline
emphasizes negotiations that result in adequate crowd expression and a reduced
potential for violence. Conceivably, group leaders may be won over to your way of
thinking. To assist tactical commanders in negotiations, this aspect of the guideline
provides suggestions for assessing points of intervention in crowd performance.

Acceptability—Redefining What it Means to Win
Winning in a crowd control situation is not like winning in combat. U.S. forces can
appear to be formidable, but they risk being portrayed as oppressors. If so, they
can lose by appearing to win. Winning in this environment is about holding the
moral high ground. This aspect of the guideline addresses acceptability of actions,
to include the use of non-lethal weapons.

Education and Training Guidelines for Crowd Control
Effective decision-making in crowd control depends on educating and training
mid- to senior-level officers. This education and training could be based on case
studies. Case studies could be used to exercise prevention planning and crisis
decision-making. An organization’s ability to learn depends on its knowledge
management—capturing lessons learned, and best and worst practices. Using
case studies not only contributes to the education and training, but also provides
feedback on tactics, techniques and procedures. Ultimately, such a system can
help an organization prepare for the challenges of crowd control.

A “Road Map” for Future Research on Crowd Behavior
No substitute exists for research. Relative to personal expertise, research provides
the most reliable information on crowd behavior. More research is needed to address
the unknowns of crowd behavior. As new and innovative theories emerge, there is an
opportunity to study traditional approaches versus these new theories to advance the
knowledge base. Such research should be designed to compare and contrast conflicting
theories of why crowds turn violent. Ideally, sufficient information would allow crowd
behavior to be modeled with fairly accurate and repeatable probabilities. To get to that
point, the Panel recommended that research be conducted on the following issues:

Why do people assemble?
Information has been collected on where and when and how people gather.


2

However, why they gather still remains to be determined. The question is important
because it may indicate the participants’ level of commitment.

What organizational structures exist within a gathering?
Research shows that a gathering consists of several clusters or smaller groups.
However, the degree of overarching organization must be studied. There may be
several levels of organization.

What determines the intensity or “energy” of a gathering?
Some gatherings seem to have greater intensity or “energy” than others. Factors
such as individual commitment, organization and preparation may contribute.

How do crowds in other cultures behave?
Little is known about crowd behavior in non-western cultures. While it is likely that
these cultures assemble, gather and disperse similar to those in the West, they
may behave differently.

What psychological and social factors contribute to violent
behavior in crowds?
Using models such as Dr. Taylor’s and other factors from the extant literature,
research should identify what psychological and social factors make a crowd prone
to violent behavior. By understanding these factors, it may be possible to predict
the likelihood of violence and intervene more effectively.

What cultural and situational factors contribute to
violent crowd behavior?
Research should examine if some cultures (i.e., religious, political) are more prone
to violent behavior and why. These findings would enable increased sensitivity to
the cues that predispose violence. Also, research should consider what situational
factors contribute to violent behavior. Several issues worthy of investigation include
crowd size, density, physical space, and resources.

What military or law enforcement actions lead to
violent crowd behavior?
Some actions of military and law enforcement have often been blamed for inciting
violent crowd behavior. Research should explore what specific actions incite
violence. This would be valuable for tactical commanders, so that their forces act
as deterrents rather than instigators of violence.

What crowd-control techniques are most effective in
preventing and stopping violent behavior?
Assessing crowd behavior and controlling crowds is a leading core capability sought
by the Joint Non-lethal Weapons Program Research should determine specific
techniques, including identifying non-lethal weapons that are most effective in
preventing and stopping violent crowd behavior, that are most effective in meeting
crowd-control objectives.


3

Introduction
Purpose
This Human Effects Advisory Panel assessed the behavior of crowds for the
purpose of determining the potential for control of crowds. Such crowd control
may encompass the effective use of non-lethal weapons. In support of this
assessment, the Panel accomplished the following objectives:
s Identification of what is known and not known about crowd behavior.
s Development of a decision-making guideline for crowd control.
s Development of a education and training guideline for crowd control.
s Development of a “road map” to guide future research on crowd behavior.

Requirements for Crowd Control
Crowd control is a leading core capability sought by the Joint Non-lethal Weapons
Directorate. This capability includes “the means to influence the behavior and
activities of a potentially hostile crowd, as well as, the capability to bring a mob
engaged in a riot under control.”1
The need to thoroughly
examine crowd behavior
Moreover, the need for this core capability continues to increase. In Somalia, U.S.
grew out of the Panel’s
Marines often faced hostile crowds of rock-throwing women and children.2 In Bosnia,
previous assessments of
U.S. Army soldiers had to disperse angry mobs of Serb hard-liners near the town of
non-lethal weapons… .
Banja Luka.3 More recently, Danish, French, and Italian forces attempted to control
riots between ethnic Albanians and Serbs in Mitrovice, Albania.4

The need to thoroughly examine crowd behavior grew out of the Panel’s previous
assessments of non-lethal weapons, which were designed for crowd control.
Weapons such as the Modular Crowd Control Munition (MCCM) and the 40mm Non-
lethal Crowd Control Cartridge (NLCDC) were developed to modify the behavior of
individuals within a crowd and ultimately the crowd as a whole. However, this behavior
modification greatly depended on an understanding of individual and group motivation.
Little is known about these subjects and therefore little could be predicted about the
probability of behavior modification using these non-lethal weapons.

This deficiency was highlighted in the HEAP’s Report of Findings: Modular Crowd
Control Munition (MCCM) and Non-lethal Crowd Dispersal Cartridge (NLCDC)
Assessment. The HEAP stated that, “data on crowd dynamics are needed for the
development and employment of crowd control devices.”5 The Panel further stated

1 NLW: Multiservice Procedures for the Employment on Nonlethal Weapons, Air Land Sea
Application Center, October 1998, pp. III-1 and III-2. Also, the Joint Non-lethal Weapons Directorate,
Joint Non-lethal Weapons Program: 1997 – A Year in Progress, February 1997, pp. 5-6.
2 Glenn Goodman, “Rubber Bullets and Sticky Foam,” Armed Forces Journal International,
February 1997, p. 26.
3 Dan Priest and Lee Hockstater, “NATO Troops, Serbs Face Off in Bosnia,” Washington Post, 9
September 1997.
4 Chris Bird, “Divided Kosovo Ignites in Riots,” The Guardian, October 16, 1999. Crowd Behavior, Crowd Control
5 Human Effects Advisory Panel, Report of Findings: Modular Crowd Control Mine (MCCM) and
Non-lethal Crowd Dispersal Cartridge (NLCDC) Assessment, 31 January 2000, p. 22 4

that, “Data should be gathered on a variety of crowd experiences” and that “A
panel of behavioral experts would then evaluate the data.”6

The Panel
The Human Effects Advisory Panel serves as an independent advisor to the Joint
Non-lethal Weapons Directorate. It reports its findings to the Directorate and
provides assessments of human effects issues relating to non-lethal weapons as
well as examines any other issues the Directorate deems appropriate. To
accomplish these assessments, each Panel is composed of recognized experts
appropriate to the assigned task.

This Panel consisted of recognized experts in crowd behavior and control. It
included a gathering of academicians, consultants and practitioners. Academicians
were selected on the basis of their research on crowd behavior. Also selected This Panel consisted of
were consultants with expertise in psychological and interpersonal dynamics of recognized experts in
individual and crowd behavior. Practitioners came from the law enforcement crowd behavior and control.
community and were chosen on the basis of their experience in crowd control and
non-lethal weapons employment. The mix of experience and expertise enabled the
Panel to address crowd behavior, finding commonality between theory and practice.

These Panel members included the following:
Previously Chief Scientist of the Occupational and Health Directorate,
Armstrong Laboratory, he is now a psychological consultant in behavior
toxicology and non-lethal technologies.
Lieutenant Sid Heal
He is a law enforcement officer in the Los Angeles Sheriffs Department. He
has extensive experience in use of non-lethal weapons and crowd control
and has been recently selected for promotion to Captain. He is Chief Warrant
Officer 5 in the U.S. Marine Corps Reserve and has participated in
peacekeeping operations in Somalia and Bosnia.
Major Steve Ijames
He currently serves in the Springfield Police Department, where he commands
its full-time tactical unit. He is a recognized expert and instructor in less-than
lethal weapons, and has published extensively on their employment.
Dr. John M. Kenny
He is an associate research engineer at Penn State’s Applied Research
Laboratory and the principal investigator for the Human Effects Advisory Panel
(HEAP). This is the sixth HEAP assessment that he has led. He was a Fellow
at the Brookings Institute and is a retired U.S. Navy Commander.
Dr. Clark McPhail
As a Fulbright Senior Research Scholar in England (1999–00) he studied
Scotland Yard videotape records on crowd behavior and riots. For more than
three decades he has studied crowd behavior and riots in the U.S. and has
published extensively on those topics. He is Professor Emeritus at the
University of Illinois.

6 Ibid., p. 21

5

Captain Dick Odenthal
Recently retired from the Los Angeles Sheriffs Department, he was
Commander of its emergency operations bureau. He has experience in over
100 riots and has advised several police departments on riot control.
Dr. Jim Taylor
He heads Alpine/Taylor Consulting, a psychology consulting firm that specializes
in individual and group behavior. He works with high-level performers in sports,
business, and police departments. Previously, he was a university professor who
published extensively on the psychology of performance.
Dr. Peter A. J. “Tank” Waddington
A professor at the University of Reading, he has written seven books on policing
as it applies to public order. He has been trained at all levels of riot control
and has seen over 100 riots. Prior to academia, he was a police officer.

Preparation for the Panel
Prior to convening, Panel members were provided with advanced reading material
on crowd behavior and control. They included the following:
s “Keeping Dissent in Its Place,” from Policing Citizens: Authority and Rights, by
Dr. Peter A. J. Waddington.
s “The Politics of Explanation and Justification,” from the Strong of the Law, Dr.
Peter A. J. Waddington.
s “Stereotypes of Crowds and Collective Behavior: Looking Backward, Looking
Forward,” by Dr. Clark McPhail.
s “Policing Protest: The Evolving Dynamics of Encounters between Collective
Actors and Police in the United States,” by Dr. Clark McPhail.
s “Policing Protest: The Control of Mass Demonstrations in Western
Democracies, Chapter 2,” Policing Protest in the United States, 1960-1995,
by Dr. Clark McPhail.
s “The Institutionalization of Protest in the United States,” by Dr. Clark McPhail.
s “The Dark Side of Purpose: Individual and Collective Violence in Riots,”
presidential address to the Midwest Sociological Society, by Dr. Clark McPhail.
s “Crowds, Mobs, and Non-lethal Weapons,” by Lieutenant Sid Heal.

The History of Hostile Crowds and Riots
Dr. Wendy Gilpin conducted background research on hostile crowds and in
preparation for the Panel’s meeting. Dr. Gilpin is a research associate at Penn
State’s Applied Research Laboratory. Her search focused on the history of this
social phenomenon. She presented her findings to the Panel. During her
presentation, Panel member Dr. Waddington provided amplifying comments
regarding how riots have changed and varied over time.

Riots have occurred in every century and every region of the world. They have
been attributed to poverty, hunger, unemployment, strikes, industrial disputes,
politics, religious and ethnic differences, sports, alcohol, and even the weather.

Riots involving poorer social classes and political groups were frequent in ancient
civilizations. An earthquake in Sparta in 464 BC prompted slaves to riot in the

6

streets. Sparta’s king led forces against the slaves. When the Roman government
abolished laws that helped the poor in 121 BC, riots erupted in the Forum and
over 3,000 people died. In 532 AD, rioting between political groups occurred in
Constantinople. The city burned for three days and 35,000 people were killed.

Riots associated with religious differences have occurred throughout history. For
example, German Protestant troops controlled Rome and rioted for days in 1527.
Catholics massacred Protestants in southern France in 1545.

Alcohol has been seen as a cause of riots. English politicians and religious leaders
argued that drinking gin encouraged laziness and crime. In 1729, Parliament passed
a Gin Act that increased the tax on the drink. This action was unpopular with many
in the working classes who rioted in response to this law in London in 1743.

Riots have been attributed to food shortages. Bread made up three-quarters of an
average person’s diet in early modern France and sudden rises in grain prices
would bring mobs to the streets. Riots occurred in Le Havre and Nantes in 1768,
in Reims in 1770, in Bordeaux in 1773. The Flour War of 1775 was a particularly
vicious example of a food shortage incited riot.

Rumors and counter-rumors reportedly have pushed mobs to action. In France in
the early 1770s, grain supplies were entrusted to a few chosen merchants, which
led to rumors that the King’s ministers had conspired to starve the King’s loyal
subjects. In early 1775, grain prices had risen over 50 percent in Paris and rioting
spread to the gates of Versailles and the streets of Paris.

Riots could be politically and religiously motivated in London during the 18th Century.
In 1780, Lord Robert Gordon, who strongly opposed Catholic emancipation, led a
crowd of 50,000 people to the House of Commons to present a petition for the
repeal of the 1778 Roman Catholic Relief Act. This crowd turned to violence and
over the next five days destroyed many Catholic chapels and private homes. It is
estimated that over 400 people were killed and £180,000 worth of property was
destroyed during the Gordon Riots.

In America, high and unfair taxation was the leading cause of many riots in the
1700s. American colonists rioted against tax collectors and other British-appointed
officials in the 1760s and ’70s. One of the most well known social protests is
associated with a tea tax. In 1770, the British government repealed all import
taxes except on tea. This tea was shipped to East India Company agents in America
who held a monopoly on its sale. Colonists feared this monopoly would put local
merchants out of business and that paying the tea tax would set a precedent for
other taxes. When tea ships arrived in Boston, colonists demanded that they be
sent back to England. The governor refused to send the ships back and colonists,
disguised as Indians, boarded the ships and dumped 342 chests of tea into the
harbor. This action became known as the Boston Tea Party.

Other American riots have occurred that have been attributed to unfair taxation.

7

Daniel Shays, a former Revolutionary Army captain, led a rebellion against
excessive property and polling taxes that prevented the poor from voting. Shay’s
rebels protested unfair court actions, the high cost of lawsuits, and the lack of a
stable currency. In 1786, these rebels stormed the courthouses in Northampton
and Springfield, Massachusetts to prevent the trial of debtors. In 1787, Shays
rebels were captured and sentenced to death for treason. They were later pardoned.

The 1794 Whiskey Rebellion was attributed to a whiskey tax. Farmers, in western
Pennsylvania and the frontier regions, found it more profitable to produce and
ship whiskey rather than bulky corn and rye crops to markets. When Pennsylvania
farmers refused to pay a whiskey tax, the federal government ordered the arrest
of the ringleaders. The farmers exchanged gunfire with government representatives,
burned the property of tax inspectors, and marched on Pittsburgh. An army of
nearly 13,000, led by George Washington, put down the rebellion.

Dr. Waddington pointed out a transformation that occurred around the end of the
18th Century and early 19th Century. Previously, most rioting was local and the
result of a direct action. However, during this transition time, the nation state vastly
expanded. Decision-making and the concentration of monetary capital shifted from
the provinces to London, and so did the form of social protest. Special interest
groups emerged and were the organizers of protest rallies and of processions
bearing petitions to Parliament. Social protests became more national, more
organized and more moderated by national organizations

This transformation was reflected in the 1819 “Peterloo Massacre” in Manchester,
England. A crowd of 60,000 working people, seeking the right to vote, gathered in
St. Peter’s Fields. These protesters were attacked by Lancashire militia. Eleven
people were killed and 400 were injured. Peterloo took on many of the
characteristics that we now see in modern protests and riots—people gathered in
central locations, carried placards, and listened to agitators.

The “Peterloo Massacre” also marked a transformation in how riots were controlled.
Previously when riots or hostile crowds formed, military forces would be called
out, a few volleys would be fired, a few people would be killed and the social
unrest would dissipate. Peterloo is remarkable for the furor that followed the action
of the Lancashire militia. The death of 11 protesters and injury to 400 others was
a prominent catalyst for the installation of a professional police force, a police
force capable of dispersing unruly crowds without resorting to excessive force.

In the United States, some 1800s riots were believed to be caused by anti-Catholic,
anti-immigrant, and anti-black sentiments. Many American citizens strongly disliked
immigrants, especially Irish Catholics and Asians. In the mid-1850s, members of
the Know-Nothing, or American Party, feared rising political power of Irish Catholics.
Irish Catholics were attacked in several cities, including Baltimore, Louisville, New
Orleans, and St. Louis.

During the American Civil War, anti-draft riots broke out in New York City and were the

8

most destructive riots in U.S. history. Armed mobs swarmed through the downtown
area to protest the draft. Rioters looted, set fire to buildings, and shot blacks, policemen,
and federal troops. More than 1,000 people were killed or wounded.

Riots associated with racial differences have been especially destructive in the
United States. Before the Civil War, violence aimed at blacks and abolitionists
occurred in northern cities. After the war, white southerners attacked blacks in
New Orleans and Memphis. Many Chinese immigrants, who were perceived as
taking jobs away from U.S. citizens and lowering wages, were victims of mob
violence during the 1870s depression.

Riots attributed to labor problems caused considerable bloodshed in the United
States. Dozens of people were killed during the great railroad strikes of 1877.
Striking workers were the targets, not the source of the bloodshed and fatalities.
The violence came at the hand of both government troops and private security
forces mobilized by industrial owners. During the 1900s, riots attributed to labor
and racial problems erupted in the United States. In 1919, efforts to unionize the
steel industry led to riots at plants in Indiana, Ohio, and Pennsylvania. In 1934, a
dispute between unions and management in the cotton-textile industry led to riots
in Georgia, South Carolina, Alabama, Rhode Island, and other states.

Violence against person and property in connection with labor protests, work slow-
downs and strikes virtually disappeared with the passage of the Wagner Act. This
act established the right of workers to strike for more favorable working conditions
and higher wages. It also recognized collective bargaining as a legitimate means
of negotiation in labor disputes.

In the early 1900s, attempts to segregate southern blacks and keep them from
voting led to lynching in rural areas and riots in cities. During World War I, many
blacks moved to the north to work in defense plants. Whites feared that blacks
would take jobs and move into their neighborhoods. Blacks claimed that white
police officers treated them unfairly. These grievances led to clashes. The worst
occurred in East St. Louis, Illinois in 1917, where 48 people died. A Chicago riot in
1919 caused 38 deaths. Racial violence broke out during World War II, the most
destructive occurring in 1943. A Detroit riot resulted in the deaths of 34 people,
and similar rioting occurred in Harlem.

Social disorder associated with ethnic hatred occurred in Europe prior to World
War II. Throughout the German Reich, a massive and coordinated attack was
conducted against Jews on November 9, 1938, also called Kristallnacht, or “The
Night of Broken Glass.” The windows of Jewish shops were smashed and store
contents destroyed. Synagogues were targeted for vandalism. This social unrest differed
from the past. As Dr. Waddington pointed out, this unrest was government inspired.
The decision to instigate it was made at the highest levels of Nazi leadership.

Riots attributed to economic deprivation and social injustices suffered by blacks in
U.S. cities erupted in the 1960s and again in the 1990s. Riots occurred in the

9

Watts section of Los Angeles in 1965, Detroit and Newark in 1967, and Cleveland
in 1968. The Detroit riot was the most violent and led to 43 deaths and property
damage of about $45 million. An unpopular verdict in highly publicized court case
in1992 sparked major unrest again in Los Angeles.

Dr. McPhail pointed out that the defining characteristic of ‘riot’ and rioting is violence
against person or property by one or more individuals in a gathering. The most
frequent riots in contemporary U.S. society are “celebration riots” that are related
to sport victories. Riots can be also be characterized as “commodity riots,” in which
most violence occurred against property rather than persons. This included
vandalism, looting, and arson. Riots characterized by violence against persons
are “communal riots,” which usually involves violence between rival racial, ethnic,
religious, or language communities.

Dr. McPhail further noted that there has been extensive violence by law enforcement
personnel against civilians, which are characterized as “police riots.” These incidents
involves police violence against civilians, who are then abandoned without arrest,
while the officers proceed to engage in violence against other civilians. The most
notable case of police rioting occurred in Chicago during the anti-war protests
during 1968 Democratic Convention.
Riots have changed with
Dr. Waddington summarized the significance of the history of hostile crowds and time and technologies.
riots. From a historical perspective, it is obvious that all riots are not the same, but Changes in social disorder
rather they are very context specific. Riots have changed with time and technologies. are likely to continue to
Changes in social disorder are likely to continue to occur and a better understanding occur and a better
of crowd behavior is imperative. understanding of crowd
behavior is imperative.
Conduct of the Panel’s Assessment
The Panel met from 12–14 September 2000 at the Renaissance Orlando Hotel in
Orlando, Florida. Convening at 1900 on 12 September, Dr. John Kenny, the Panel’s
principal investigator, reviewed the Panel’s agenda, objectives, and procedures.
Dr. Don Farrer was elected as the Panel’s chairperson.

Amplifying instructions for the Panel’s assessment of crowd behavior were provided
by Major Noel Montgomery USAF of the Joint Non-lethal Weapons Directorate.
He gave an overview of the Joint Non-lethal Weapons Program, which included
the Directorate’s need to understand crowd behavior.

Dr. Wendy Gilpin of Pennsylvania State University presented her research on past
riots and hostile crowds. She was followed by Major Steve Simpson’s presentation
on the Defense Department’s current crowd control techniques, tools and training.
The Panel then adjourned at 2200.

The Panel reconvened at 0800, 13 September. Dr. Jim Taylor was the day’s first
speaker and described individual and crowd behavior models, with emphasis on:
key questions to be raised; behavior modification; and intervention points for police
and military forces. The Panel then received a briefing on policing actions and the

10

use of non-lethal weapons from Major Steve Ijames.

Dr. Clark McPhail briefed the Panel on purposive action across the life course of
temporary gatherings and the rioting that sometimes develops from those
gatherings. His presentation was followed by Lieutenant Sid Heal, who briefed the
Los Angeles Sheriffs Department’s heuristic approach to crowd control and riots.

Captain Dick Odenthal briefed the Panel on the strategic and tactical considerations
for crowd control. The last brief of the day was presented by Dr. P.A.J.“Tank”
Waddington, who covered the political implications of policing public order.

The Panel adjourned on 14 September at 1400. The findings and recommendations
of the Panel are presented in the following sections of this report.


11

Objective 1
WHAT IS KNOWN AND NOT KNOWN ABOUT
CROWD BEHAVIOR

General
Many military and law enforcement concepts on crowd control are based on
stereotypes that have been disproved by social scientists. The findings of these
scientists indicate a need to reevaluate existing methods and operating procedures
for crowd control. Gatherings are not the entity that the term “crowd” often implies.
Rather, gatherings are more of a process with a beginning, middle and end. While Many military and law
such gatherings may lend themselves to rioting, they seldom result in violent activity. enforcement concepts on
Additionally, many of the common beliefs regarding what causes riots are not crowd control are based on
supported by empirical evidence. stereotypes that have been
disproved by social scientists.
Discussion
Many of today’s views on crowd behavior are based on stereotypes that grew out
of academia over a century ago. Dr. Clark McPhail pointed out that the writings of
Gustave LeBon in 1895, Robert Park in 1904 and 1930 and Herbert Blumer in
1939 long dominated scholarly thinking about crowd behavior. Their writings
fostered the beliefs that crowds are unified masses whose behaviors can be
categorized as active, expressive, acquisitive or hostile. Their writings also held
that crowd participants were given to spontaneity, irrationality, loss of self-control,
and a sense of anonymity. Other scholars proposed “that crowds do what they do
because of their lower class, criminal or riff-raff composition.”7

Debunking Stereotypes
These stereotypes were initially repudiated by sociologist, Dr. Carl Couch in 1968.8
In the three decades since Couch’s overturn of these stereotypes, extensive Crowd participants are
research by many other social scientists have consistently affirmed Couch’s critique not unanimous in their
and rejection of these stereotypes. Essentially, these scientists found that: motivation.
s Crowds are not homogeneous entities – all the participants are not the same.
s Crowds are not made up of isolated individuals, but of a minority of
individuals and a majority of small groups of people who are acquainted
with one another.
s Crowd participants are not unanimous in their motivation.
s Crowd participants are not anonymous to one another.
s Crowd members are not given to unique or distinctive emotional displays.
s Crowd participants seldom act in unison, and if they do, that action does
not last long.
s Crowds do not cripple individual cognition.
s Crowds are not uniquely distinguished by violence or disorderly actions.

7 Clark McPhail, “Stereotypes of Crowds and Collective Behavior: Looking Backward, Looking Crowd Behavior, Crowd Control
Forward,” Studies in Symbolic Interaction, Supplement 3, 1997, p. 37.
8 Ibid. 12

s Socioeconomic, demographic and political variables are poor predictors of
riot intensity and individual participation.
s Individual attitudes and personality characteristics are poor indicators of riot
participation.

These findings are well known among scientists who have studied collective behavior.
The problem, though, as McPhail pointed out, is that they have not been reflected in
introductory sociology and psychology textbooks that still contain many of the
stereotypes regarding crowd behavior. Students taking advanced sociology and
psychology courses get a more accurate understanding of crowd behavior. However, Socioeconomic, demographic
students that do not take more than an introductory course carry these crowd and political variables are
stereotypes into such careers as journalism, law enforcement, the military, and public poor predictors of riot
policy.9 intensity and individual
participation.
These stereotypes have also found their way into military and law enforcement concepts
of crowd control. Dr. McPhail has found such stereotypes in his review of civilian and
military publications on crowd control. Moreover, stereotypes are seen in the Non-
lethal Individual Weapons Instructors Course, which is based on law enforcement
training programs. Its performance steps call for military personnel to determine the
“crowd type.”10 It is indicative of the misleading crowd types that Herbert Blumer
advanced in 1939. It also implies unanimity of motivation and action within a crowd.

The Crowd: A Process—Not An Entity
Empirical evidence indicates that defense and law enforcement agencies should rethink
their views of crowds. It shows that crowds are not the entities that some have implied.
Rather, crowds are a gathering of a multitude of individuals and small groups that
have temporarily assembled. The small groups are usually comprised of friends, family
members, or acquaintances. Individuals do not assume a sense of anonymity in a …crowds are a gathering of
crowd. Individuals are known to companions in their group, and to others in the gathering a multitude of individuals
that have come from the same neighborhood or community. and small groups that have
temporarily assembled.
Such gatherings are really more of a process. They have a beginning, middle and an
ending phase. This process is reflected in the chart below.

Figure 1.
The phases of a gathering. Gatherings have three phases; an assembling process,
the temporary gathering, and a dispersing process.

9 Ibid. Crowd Behavior, Crowd Control
10 Major Steve Simpson, briefing entitled “The US Department of Defense Joint Non-lethal Weapons
Program: Training,” 12/09/00, slide 10 “Crowd Dynamics/Crowd Control.” 13

Military and law enforcement personnel focus on the end phase—dispersal. However,
they must also consider how people assemble and interact during the course of the
gathering. These aspects of a gathering are likely to affect how it disperses.

The Assembling Process
This refers to the movement of people from different locations to a common location,
within a given period of time. It largely determines who participates in a gathering.
Examination of complex protest actions shows that such collective phenomena
do not emerge randomly or spontaneously. Rather they are the result of prior
activities by purposive actors, and the more complex the collective action, the
more likely the prior planning and organization.

Assembling can occur in one of two ways: on an impromptu basis, or by an
organized mobilization effort. Both involve prospective participants telling one
another what is happening, when and where, and a suggestion or invitation to
participate. The big difference between the two processes is that organized
mobilization efforts emanate from one common source. Such mobilization efforts
tend to provide justification. For example, they might say, “we have to stand up for
our rights which are under attack,” or “we know you are one of us and we need
your help.” Organized mobilization relies heavily on established networks that attract
people to a gathering en bloc. This was seen in the recent anti-capitalism/anti-
globalization protests in Seattle and elsewhere that mobilized far-left radicals,
anarchists, environmentalists, campaigners against Third World debt, labor unions
and many more—all with slightly different and potentially conflicting agendas.

The impromptu process usually develops informally, and is mostly done by word
of mouth. It does not always include justification. However, it can be aided by the
media, as was the case in the riots that occurred after the court decision on the
Rodney King beating. Several studies have examined the impromptu process. Dr.
Benjamin Singer studied how participants learned about the 1967 Detroit riot. His
data showed that most participants learned through interpersonal contacts, either
by telephone call, or someone knocking on the door. He also found that people,
who participated immediately, were already close to the riot. Those that were in
more remote locations did not rush to the riot because, in many cases, they were
already engaged in other routine tasks and obligations.

Dr McPhail also studied the impromptu assembling process in a 1969 celebration
riot after a college basketball upset victory. Fans gathered at the Columbia, South
Carolina airport to greet the University of South Carolina basketball team that
returned shortly after the game. Data showed that participants learned about the
gathering from friends and acquaintances. The best predictor of who received
assembling instructions was prior basketball fan activity. The best predictor of who
assembled was the frequency with which instructions were received. The factors
that facilitated assembly were access and availability. Par ticipants had
transportation and no competing demands.

The mobilization process relies more on centralized planning and organization.

14

Organizers have lists of names, and they contact potential participants through
contact networks, telephone banks, mailing lists or e-mail addresses. They also
attempt to schedule the event so that it does not conflict with other major events.
Additionally, they may arrange transportation for participants. Several studies have
found these to be common factors in the mobilization processes for political protests,
Billy Graham crusades, as well as, presidential and senatorial campaign rallies.

The Temporary Gathering
Temporary gathering refers to the collection of individuals and small groups in a
common location. It is the result of the assembling process. The temporary gathering
consists of the individual and collective actions.

Studies have determined that two or more people can regularly engage in about
forty elementary forms of collective action. All of these actions do not occur in
every gathering. They represent the range of actions that have been observed in a
wide variety of political, sport, religious and leisure gatherings. These data are the
result of more than three decades of direct observations of hundreds of gatherings.

To determine the degree of collective behavior, researchers matched this list of 40
actions against their observations of the March for Life demonstration that occurs
each year on 21 January in Washington DC. Researchers were placed at the …uniform actions in
assembly area, along the march route and at its destination, the Supreme Court. gatherings seldom occur.
The study found that there were no collective actions in which all participants
engaged at the same time. On occasions, 50 to 75 percent of all participants
engaged in standing together, or facing in the same direction. Less than 50 percent
of the people engaged in carrying placards and banners at one time. Less than 50
percent were engaged in shouting, praying or chanting at one time.

There is a much greater participation in collective action in religious gatherings.
This may be because there is an order of worship for participants to follow, in
effect a common script. Even so, the proportion of observed worshippers engaging
in collective acts is between 80 and 90 percent. It seldom reaches 100 percent.

These studies question the illusion of unanimity in gatherings. They show that
uniform actions in gatherings seldom occur. When they do occur, these actions
are usually very simple and they do not last long. For example, spectators at a
football game will collectively moan failures and cheer successes of their team.
Afterwards, individuals will return to a variety of individual actions such as talking
with companions, reading programs, eating or drinking.

Such studies also provide familiarity with the repertoire of collective actions within
a given culture. It is important that U.S. forces know these repertoires of actions
when they operate in different cultures. Dr. Waddington provided the example of
the Zulu tribe in Africa. At some point in a ‘political’ gathering, they may sprinkle
water on one another. If so, this is an indication they are preparing to attack. Cultural
anthropologists may provide information on such repertoires of collective action in
non-western cultures.

15

Studies have shown that people engage in “purposive actions” in gatherings. This
means that, although they may not hear or see everything, they are in control of
the information they process. In other words, they have not subordinated or
surrendered their ability to reason to the crowd or its leaders. If they engage in
collective behavior, it is because they have chosen to do so.

How they decide to engage in collective behavior varies. They may independently
engage in collective actions based on some recognizable cue. For example, when a
performance ends, most people recognize that it is time to applaud. They also may
interdependently decide to engage in collective behavior. For example, two or more
companions may consult with each other before deciding to participate in some activity.
Additionally, two or more people may adopt the goal of a third party. This may occur
when the police order a gathering to disperse and people choose to follow that order.
…people have not
subordinated or
The Dispersing Process surrendered their ability
This is the last phase in the life of a temporary gathering. It involves the movement
to reason to the crowd or
of people from a common location to one or more alternate locations. Dispersal
its leaders.
brings the temporary gathering to an end, or at least begins its decline. Dispersal
can occur on a routine, emergency and coerced basis.

The routine dispersal may be specified in advance. It also may even be part of the
assembling instructions. Such instructions allow participants to plan for routine
personal needs – working, shopping, eating, sleeping, etc. Routine personal needs
can even affect participation in protest events and riots. This was seen in the riots
in Watts (1965), Detroit (1967), Washington DC (1968), and Miami (1982).
Researchers plotted the number of violent events occurring in each riot (see Figure
2). They found that riot events decreased at times when people ordinarily engaged
in such personal activities as working, eating, or sleeping. Researchers have also
found that most riots started after working hours and on weekends.


16

Figure 2.
Distribution of Riot Events by Time of Day for the Watts (1965), Detroit (1967), and
Washington DC (1968)

Lieutenant Sid Heal of the Los Angeles Sheriffs Department also commented on
this trend. He participated in a law enforcement workload study that plotted 300,000
calls for service in riots, fires and earthquakes. He stated that these calls for service,
after being plotted, closely paralleled the trends found in the four riots previously
mentioned.

There may be exceptions to this trend. Captain Dick Odenthal, who recently retired
from the Los Angeles Sheriffs Department, found that the 1992 Los Angeles riots did
not follow previous trends. He stated that a large number of riot events occurred when
least expected, for example, in the mornings, between 9:00 AM and 12:00 noon.

Emergency dispersal occurs when people evacuate an area as the result of some
unexpected crisis, such as a fire, explosion, or flood. Several stereotypes such as
incapacitating fear, mass panic and irrational behavior are associated with
emergency dispersal, and they have been perpetuated by mass media. However,
these stereotypes have been refuted by extensive research on the behavior of
individuals and organizations in disasters over the past half-century.

During the Eisenhower administration, the federal government funded research
on how civilians, as well as fire, medical and public safety personnel, might behave
during a nuclear attack. That research documented that those personnel were not
distracted, individually or collectively, from their professional responsibilities during
tornadoes, floods, and other natural disasters. Sociologists have also studied
individual and small group collective behavior in crisis events like the fire at the
Beverly Hills Supper Club near Cincinnati, Ohio. They found that individuals in
such emergency situations quickly recover from initial shock. They keep their wits


17

about them and improvise ways of extricating themselves and their companions from
dangerous situations. Additionally, research found that most people give as much
effort and attention to others’ well-being, as they do their own personal safety.

In some emergency situations, tragedies have occurred that have been mistakenly
attributed to panic. This stereotype holds that after an explosion or fire people exit
theaters or nightclubs in panic, and mindlessly trample others to death. However,
the reality is different. Carefully conducted interviews with survivors have found
that people towards the rear of a gathering could not see or hear what was occurring
in front of them. As a result, they mistakenly continued to move toward a blocked
exit and tripped over fallen individuals. Most people that are killed in these situations
are asphyxiated beneath a mass of fallen bodies.

Those forces on the scene must also be careful not to misinterpret what is occurring
in an emergency dispersal, as Dr. Waddington pointed out. He stated that this
occurred when soccer fans were mistakenly believed to have become disorderly
at a game in Sheffield, England. In an effort to get out of the stadium and avoid the
exiting crush, many bystanders attempted to climb fences. The police mistook
…negotiated management
them for soccer hooligans and began hitting the fence with batons. As result, these
of crowds has increasingly
individuals fell back into the crowd. and died from crushing.
become the preferred method
of getting protest organizers
Dispersal may also occur as the result of coercion. Prior to the 1970s, law
to police themselves.
enforcement often used escalating force to disperse gatherings. They assumed
that crowd members were irrational or victims of a crippled crowd mentality. Thus
police resorted to the use of force to move people who refused to disperse. This
began to change during the anti-war riots in the late 1960s and early ’70s. The
Washington DC police realized the benefits of contacting and negotiating with
demonstration organizers.

Since then, negotiated management of crowds has increasingly become the
preferred method of getting protest organizers to police themselves. Lieutenant
Heal and Captain Odenthal reinforced this point, indicating that negotiated
management has been highly successful when demonstration leaders were
available and willing to participate.

What We Know About Riots
A riot is judged to occur when one or more individuals within a gathering engage
in violence against person or property. The 1965 Watts Riot, the 1980 Miami Riot
and the 1992 Los Angles Riot exemplify such violence.

Frequency of Riots
Most demonstrations and protests in the U.S. and Europe have not resulted in
riots. Protest can evolve into violence, as was the case in the 1968 Columbia,
South Carolina civil rights protest, the 1969 Weathermen’s Days of Rage in Chicago,
and the 1992 Poll Tax protest in London. However, based on empirical and
quantitative records on protest demonstrations in Europe from 1830 to 1930 and
the U.S. records from the 1960s to the present, less than 10 percent of protest

18

demonstrations have involved violence against person or property. Little data have
been gathered on other cultures.

As mentioned previously, the most frequent kind of riot in recent U.S. history has
been the celebration riot. Such riots occurred after the Chicago Bulls NBA
championships in 1992, the Denver Broncos Super Bowl victory in 1999, the
Michigan State NCAA championship in 1999 and Los Angeles Lakers NBA
championship in 2000.

Community Demographics …less than 10 percent of
Several common beliefs exist regarding why riots start. Many assume that a protest demonstrations have
community’s characteristics contribute to riots. The assumption is that communities involved violence against
with significant socioeconomic, political or demographic strains will experience person or property.
riots—and the more severe the strain, the more intense the riots.

The empirical evidence does not support this premise. Dr. S. Spilerman studied
the issue of community background as it related to riots and their intensity. He
looked at the characteristics of 262 U.S. cities in which riots occurred. He also
looked at several hundred other cities with populations of 25,000 and greater, and
with the same socio-political attributes. Additionally, he had a good measure of
the intensity of the rioting in these cities. Some cities had more than one riot.
Overall, Spilerman found that no socioeconomic or political attribute of communities
consistently predicted the intensity of these riots.11 His work is regarded as the
best social science analysis that can be done with aggregate or census tract data. The empirical evidence does
not support… The
Some studies have examined those demographically similar communities that assumption is that
experienced riots and those that did not. Dr. Donald Warren conducted a study of communities with
neighborhoods in Detroit before the 1968 riots, and then returned after the riots to significant socioeconomic,
learn which neighborhoods had been involved. He found that neighborhoods with political or demographic
a high degree of public participation did not experience riots. This participation strains will experience
included involvement in neighborhood associations, private ownership of riots… .
businesses, and local representation on the city council. Such neighborhoods were
also more likely to have people try to stop rioting when it occurred.

Participants
Many sociologist and political scientists initially embraced the hypothesis referred
to as D-F-A–deprivation, frustration and aggression. They believed that individuals
that were politically and/or economically deprived were more likely to aggress to
reduce the frustration resulting from their deprivation. Most survey research on
individual riot participation in the 1960s and 1970s tested this hypothesis. Dr.
McPhail conducted secondary statistical analyses on 12 such studies of individual
riot participation in Watts, Newark, Detroit, Omaha, and Milwaukee.12 He looked

11 Seymour Spilerman. 1976. “Structural characteristics of cities and the severity of racial
disorders.” American Sociological Review. Vol. 41:771-793. The one exception to this was the
low correlation between the percentage of African-Americans in Southern cities and the intensity
of rioting. This is subject to different interpretations ranging from greater anger on the part of
Black residents to the greater suppression of Black residents by White police officers. Based on
other studies, the initiation of rioting and the subsequent pattern of development is more
frequently the result of the interaction between civilians and police than it is solely or primarily the
result of the actions or attributes of either category of actor.
12 Clark McPhail. 1971. “Civil disorder-participation: a critical examination of recent research.” Crowd Behavior, Crowd Control
American Sociological Review. 36:1059-1073. At the time the original research was done, it was
common practice to report only levels of statistically significant differences between rioters and 19

at the comparisons of samples of participants in these riots and samples of those
in the same community that did not participate. He found no support for the
deprivation, frustration, and aggression hypothesis. Extensive experimental research
was also done in the 1970s and ’80s on the D-F-A hypothesis. No connection was
found between deprivation, frustration and aggression hypothesis. Individuals that
were politically and/or economically deprived were not more likely to aggress to reduce
the frustration resulting from their deprivation. Moreover, some experimenters reported
that aggression increased rather than decreased frustration.13

Extensive research was done on participants in the civil rights movements in the Individuals that were
1960s and ’70s. When the civil rights movement began, many assumed that the politically and/or
people who were the most deprived would participate. In reality, upper middle- economically deprived were
class individuals were the initial participants in the civil rights demonstrations. The not more likely to aggress to
social networks and organizations, to which they belonged, were better predictors reduce the frustration
of participation than were any socioeconomic or demographic characteristics.14 resulting from their
deprivation.
Research was also done on participants and non-participants in the anti-war
movement. Some researchers looked at the personality traits of participants at
the University of California–Berkley and University of Wisconsin–Madison.15 They
found no evidence to support the hypothesis that any personality type or trait
predicted participation.

Considerable study was also done on who was participating in the anti-war
movement. It looked at socio-economic, political and demographic characteristics
and the personality traits of those at Berkley and at Madison. These studies also
did not find support for the deprivation, frustration and aggression hypothesis.

It is also commonly believed that some “spark” sets off riots. In the U.S., it is called
a “precipitating incident.” In Great Britain, it is called the “flashpoint.” However,
such incidents are inferred as the “spark” after the fact. Dr. Waddington stated that
he monitored about 100 public order operations. In each, the intensity went up
and down and multiple incidents occurred that seemingly could have precipitated
a riot. When an incident was followed by a riot, it became easy to say that it was
the “spark” after the fact.

Captain Odenthal made a similar observation. He stated that many people saw
the court decision on the Rodney King beating as the precipitating incident for the

non-rioters with characteristic “x”. This became a very controversial methodological issue, which
was eventually resolved by computing the extent or magnitude of correlation between
characteristic “x” and participation for both rioters and non-rioters as well as the level of statistical
significance. Dr. McPhail was the first to compute both levels of statistical significance and the
magnitude of correlation between riot participation and the various measures of deprivation and
frustration that other researchers had collected.
13 Leonard Berkowitz. 1969. Roots of Aggression: A Re-examination of the Frustration Aggression
Hypothesis.” New York: Atherton Press. Michael Quanty. 1976. “Aggression catharsis:
experimental investigations and implications.” In Russell Geen and E. C. O’Neal (Eds)
Perspectives on Aggression. New York: Academic Press. Brush, Stephen. 1996. “Dynamics of
theory change in the social sciences: relative deprivation and collective violence.” Journal of
Conflict Resolution. 4-: 523—545.
14 Maurice Pinard, Jerome Kirk and Donald von Eschen. 1969. “Processes of recruitment in the
Sit-in movement.” Public Opinion Quarterly. Vol. 33:355-369. Donald von Eschen, Jerome Kirk
and Maurince Pinard. 1971 “The organizational substructure of disorderly politics.” Social
Forces Vol. 49: 529-544. Crowd Behavior, Crowd Control
15 William Aron. “Student activism of the 1960s revisited: a multivariate analysis.” Social Forces.
Vol. 52, 1974, pp. 406-414. 20

1992 Los Angeles riots. In reality, there were multiple incidents that seemingly
could have precipitated these riots. The problem is that a precipitating incident is
usually hypothesized after the event. It has little utility in determining a course of
action for controlling crowds.

Personal experience may be the best means of forecasting potential riots. During
their presentations, Lieutenant Heal and Captain Odenthal stated they had correctly
forecasted the possibility of a riot in a given area. Dr. Waddington stated that he
observed the same forecast accuracy with police in Great Britain. They often
perceived that a particular incident in a given area might result in community …a precipitating incident
violence, and therefore initiated emergency plans and call-ups of police reserves. is usually hypothesized
after the event. It has little
The ability to make such assessments was qualified by Dr. Waddington. He stated utility in determining a
that the richness of experience enables individual officers to make such course of action for
assessments. They can sense the ebb and flow of events in a familiar community controlling crowds.
and pick up cues. They also receive intelligence, which they match to their
perceptions. As a result, such officers can intuitively assess a situation and very
likely they will be correct. As Dr. Waddington noted, there is no substitute for this
experience when making such assessments.

The problem is establishing criteria for broader social behavior. While personal
experience is undoubtedly an invaluable asset, there is, as yet, no systematic
evidence on its utility across time. Dr. Waddington pointed out, that social scientists
are obliged to be systematic, and reduce their understanding of behavior to a
checklist that shows what will and will not be effective in understanding and acting
in crowd situations. They look at all the information that pours in and attempt to …a great deal of what is
make sense of it. Social scientists do that with crowd behavior data. assumed to be true about
crowds has not been
Conclusion supported by empirical
“What we know is a very integral part of our planning” stated Major Steve Ijames. evidence. …military and
However, a great deal of what is assumed to be true about crowds has not been law enforcement must
supported by empirical evidence. This does not necessarily invalidate the incorporate the most
experiences and perceptions of military and police personnel. However, this relevant scientific findings
divergence in perspective needs to be considered and reconciled in the future. At into their current thinking
a minimum, military and law enforcement must incorporate the most relevant and modify their
scientific findings into their current thinking and modify their understanding of crowd understanding of crowd
action and their tactical approach accordingly. action and their tactical
approach accordingly.


21

Objective 2
DECISION-MAKING GUIDELINE FOR CROWD CONTROL

General
A new approach to crowd control is needed. To facilitate this new approach, the Panel
developed a decision-making guideline. This guideline emphasizes prevention rather
than confrontation. It makes tactical commanders aware of key information about crowd
encounters. Additionally, it encourages relationships with group leaders to minimize A new approach to crowd
confrontation. Lastly, this guideline addresses public acceptability of crowd control control is needed.
practices, particularly as it relates to the use of non-lethal weapons. …emphasizes prevention
rather than confrontation.
Discussion
The Department of Defense should re-conceptualize crowd control. Its current
approach is based on questionable stereotypes. What is known about crowd
behavior and crowd control indicates that the military should emphasize prevention
rather than confrontation. Crowd control requires its own thought process. In
combat, military forces fight and eliminate opponents. In crowd control, military
forces must deal with non-combatants that have internationally recognized rights.
These rights must be respected while maintaining public order. This is an issue
with which the police have been struggling for years.

This new approach requires a decision-making guideline. As Dr Waddington stated,
“dealing with crowd control incidents is a matter of playing by some basic precepts.
It also entails asking the right questions and in a logical manner; otherwise, key
issues will be missed.” The Panel, therefore, developed such a decision-making
guideline. It does not offer answers. Rather, it offers the precepts and key questions
which practitioners and academicians believe are basic to crowd control. This
guideline consists of the following:
Encouraging Relationships and Ownership
Acceptability—Redefining What it Means to Win

“You don’t write poetry in a gunfight,” as Captain Odenthal states. “If you make it
up as you go, you lose.” This phase of planning begins well before the incident and
is initiated above at the strategic level. It includes strategic guidance on crowd
control as well as addressing responsibilities, training, organization, operating
procedures, and rules of engagement. Major Ijames points out in his writings, “the
most difficult decisions for management in a tactical environment—and the most
productive—are those made prior to the event taking place.”


22

AVOID CONFRONTATION
“It is a game you don’t want to play unless you have to, is the message the police
tell me,” states Dr. Waddington. Crowd situations are highly unpredictable. However,
one thing seems certain. Confrontation will likely cause crowd resistance. Law
enforcement officers have witnessed it, and social scientists call it the “closed-
loop, negative feedback” model. Essentially, purposive human beings—civilians
or police—tend to resist opposition to the realization of their purposes.

FOCUS ON PREVENTION
Planning should key on the prevention of bad outcomes. As research indicates, Confrontation will likely
most crowds don’t turn violent. Many law enforcement agencies, therefore, attempt cause crowd resistance.
to help crowds accomplish their goals within the law. They even concede some
violations for the sake of avoiding confrontation. However, they maintain a law
enforcement presence, which signifies social restraint. These agencies have also
made decisions regarding when and where they will not compromise, or “die in a
ditch” as Dr. Waddington stated.

Such planning with a focus on prevention has been successful. Captain Odenthal
estimated that the Los Angeles Sheriffs Department’s emphasis on prevention
has succeeded more than 90 percent of the time. This is based on a five-year Planning should key on the
observation period, with an average of five demonstrations a month that had the prevention of bad outcomes.
potential for bad outcomes.

ABOVE ALL—WHAT ARE YOU TRYING TO ACCOMPLISH?
Planning should also focused on – “what are we trying to accomplish?” It the most
fundamental question that must be answered, states Dr. Jim Taylor. However,
“keeping the main thing, the main thing” is easier stated than done. Process can
sometimes drive the mission, as Major Ijames stated. This can occur in any
organization. It can easily occur in organizations that are trained to fight and
eliminate opponents, unless they focus on what they are trying to accomplish.

…the assessment of a
Practitioners on the Panel stated that they seek to know as much as they can
group’s goals compared to
about social protest groups well in advance of an incident. It not only provides
the goals of U.S. forces helps
to understand and avoid
warning, but insights as well. As Dr. Waddington stated, when you know as much
as you can about these organizations “… you can start to read the situation through
potential conflicts.
their eyes …you can anticipate what they are going to do.”

CONTINUALLY ASSESS
“What we are trying to accomplish” must be continually assessed against what
others are trying to accomplish. Some groups may not have goals that conflict
with those of U.S. forces. However, while their goals may not conflict their actions
might. Other groups may have goals and activities that do conflict. In either case,
the assessment of a group’s goals compared to the goals of U.S. forces helps to
understand and avoid potential conflicts.


23

ASK KEY QUESTIONS ABOUT THE CROWD
Assessing crowds has its own set of questions. They are based on what social
scientists and practitioners know about crowd behavior. These questions should
be answered well before a crowd assembles, or as quickly as possible if a gathering
occurs without notification. In such cases, all the questions may not be answered.
However, these questions represent a logical way to think about crowds. These
questions are as follows:
s Who are they? What is the overarching identity of the crowd? Are they strikers,
ethnic factions, or social protestors? Understanding who they are will indicate
what they may do. Do they identify themselves as strikers, ethnic groups,
religious factions, or protestors against some perceived social injustice? It
may be possible to determine identities and goals from advance assembling
instructions, leaflets distributed to bystanders, placards and banners being
carried, as well as, chants and songs being expressed. Dr. Waddington related
a negative example. During an animal rights demonstration in London, police
were prepared to protect a fur store against possible attack. However, the
What the group wants to
demonstrators stormed a pharmacy instead. Why? The demonstrators were
accomplish by assembling
protesting animal experimentation.
could determine the extent
s What are their goals? What the group wants to accomplish by assembling
to which they can be
could determine the extent to which they can be accommodated once they
accommodated once they
have assembled. They may seek only recognition for their cause. This usually
have assembled.
means being seen and heard. If so, this goal is usually easy to accommodate.
Practitioners on the Panel related how they facilitated marches for a variety of
organizations. In some cases groups may have more demanding goals. For
example, demonstrators in Seattle sought to stop the meeting of the World
Trade Organization. Such goals cannot be accommodated by tactical
commanders, making confrontation likely.
s What is the composition of the crowd and are there any known factions? As
previously indicated, crowds and demonstrations are frequently more
heterogeneous than homogeneous. The Seattle demonstrations against the
World Trade Organization were comprised of groups that were protesting
environmental issues, worker wages and child labor laws. Additionally, differing
goals and the resulting friction were evident between such organizations as
the Ruckus Society and the AFL-CIO. Factions within a crowd represent threats
and opportunities
s What are they capable of doing? Protest groups often claim that they will
assemble large numbers of people to produce some disruptive action. However,
there is often a difference between the claim and reality. One organization
claimed that it could mass 100,000 demonstrators. In reality only 40,000
showed. Dr. McPhail notes that organizers almost always claim that they can
mobilize more people than they actually do. Organizers exaggerate for two
reasons. First, they want to boost their own people’s morale, and second they
want the media to report that they have strength in numbers. Studying a group’s
past activities may provide an indicator of what they are capable of doing in
the future.
s What are their traditional behaviors or cultural repertoires? What people do
during protests is not universal. It varies with the group as well as the culture.


24

For example, soccer hooligans do not throw “petrol bombs.” Other groups
may. Social protest organizations and striking unions will carry placards and
banners. Major Ijames recounted the standard behavior of 5,000 women in
Sarejevo every other Sunday. They sat and blocked traffic to protest the loss of
male relatives in the Bosnian war. Understanding the goals of the protestors can
be helpful in deciding how to respond to their behaviors.
s When and where will they assemble? Every protest organizer has a time and
place for assembling and perhaps a destination for the crowd to move. If the
organizer attempts to mobilize large numbers of participants, the time and place(s)
for assembling and dispersing must be made known in the instructions. This
information may be stated in mobilization instructions, or disseminated by an
informal network, such as word of mouth. Such impromptu networks in densely
populated areas can enable rapid assembling. Major Ijames related an instance
in Haiti where people rapidly massed in and on vehicles outside the U.S. embassy.
s Where will they go? Many crowds have destinations. Organizations may march a
specified distance to ensure that that their cause gets sufficient attention. If so,
tactical commanders should know the route and minimize disruptions to the rest How dispersal occurs must
of society. In Los Angeles, law enforcement agencies attempt to reroute traffic be considered.
and prevent congestion caused by protest marches. They also provide security
along the route. Dr. McPhail reported that in Denver and Washington, DC, police
provided protective escorts to the Ku Klux Klan during their movement from one
point to another. The absence of protection would have assured confrontation
with counter-demonstrators and created more disorder in the community than
did the Klan’s protected movement.
s What are the possible targets of violence? Riots in particular may focus on target
facilities. In the 1992 Los Angeles riots, gun stores were major targets. In developing
countries, targets may be more basic. In Somalia, throngs of people stormed
food supplies. In Haiti, riots broke out over trash dumps where U.S. forces had
discarded the remnants of MREs (Meals Ready to Eat).
s What is the worst case scenario? The worst case scenario must be recognized
for the sake of avoidance. For many tactical commanders, this may be the situation
of a peaceful crowd degenerating into a violent riot. Dr. Waddington provided an
example of a worst case scenario. During British crowd control operations in
Belfast, a ten-year old girl was killed by the impact of a non-lethal baton round.
Nationalist propaganda flyers have shown an image of her in a coffin, which has
had a profound impact on public opinion.
s When and where will they disperse? Crowds have a life course, as Dr. McPhail
states. How dispersal occurs must be considered. Lieutenant Heal related the
example of a 70,000-man march that went from east to central Los Angeles.
Upon reaching its destination, the march engaged in peaceful activities and then
dispersed. The law enforcement agencies provided transportation back to the
origin of the march; otherwise, the protestors would have returned to their point of
origin in a disorganized and unsupervised manner. It is also essential that there
be one or more avenues which people can use to disperse and these dispersal
routes should be clearly defined, visible and open-ended avenues.
s Are there plans for subsequent gatherings? A crowd may disperse for a short
time so that participants can take care of personal needs. This may be only


25

for a few hours or it could be days. Plans for subsequent gatherings may be
part of the mobilization instructions or they may be issued at a later time.

CONSIDER POINTS OF INTERVENTION FOR INDIVIDUALS
During the life course of a crowd, tactical commanders may have to stop actions
that threatens property, personal safety and public order. As Lieutenant Heal noted,
if such behavior is not stopped, it may create a sense of empowerment and
encourage more destructive behavior.

The problem is how to stop destructive behavior. Some police often see violent …a few individuals usually
individuals in crowds, but generalize this behavior to the entire crowd. Dr. McPhail engage in destructive
states that such generalizations are usually incorrect and dangerous. Dr. behavior. The rest of the
Waddington pointed out that only a few individuals usually engage in destructive crowd is not involved.
behavior. The rest of the crowd is not involved. This observation was based on
McPhail and Waddington’s detailed analysis of videotapes of the 1990 Poll Tax
protest in London and other protests as well. The danger occurs when police
respond collectively to the entire gathering and provoke individuals who were not
involved in the violence. If this happens, their complaint is likely to be, “We weren’t
doing anything and the police attacked us,” as Captain Odenthal stated.

Police have found it more effective to target the individual(s) engaged in violent
behavior and to respond to that behavior, rather than the crowd as whole. Such
intervention requires an understanding of how this behavior might be altered. Dr.
Taylor developed a model comprised of five psychological factors for understanding
and assessing individual behavior. This model forms a pyramid, and the factors Police have found it
closer to the top have the more immediate influence on performance. more effective to target the
individual(s) engaged in
violent behavior and to
respond to that behavior,
rather than the crowd
as whole.

Figure 3.
Potential points of intervention for control of individuals.

26

Each psychological factor represents a possible point of intervention. These factors
are addressed in ascending order on the pyramid:
s Motivation – An individual’s ability to be committed to the crowd’s cause despite
fear, fatigue, opposition, and personal needs. Within a crowd, individual
motivations vary. They may be zealotry, fear, self-protection, obedience, or
selfish needs. These motivations relate to goals – what the individual hopes
to accomplish. Tactical commanders may satisfy motivations through resolution.
They also may reduce them through fatigue, fear, and pain.
s Confidence – An individual’s belief in his ability to accomplish goals. Confidence
is key to the use of that ability, and the single most important aspect of individual
behavior. Confidence enables motivation, the ability to handle stress, focus
on a goal, and the ability to experience positive emotions. Preparation,
resources, and initial success contribute to confidence. Conversely, the denial
of these can reduce confidence.
s Stress – Physical, mental, or emotional reactions to internal or external Working relationships
demands. This is largely how individuals perceive and respond to stimuli. Some between tactical
see difficulty and shrink from it. Others see challenge and respond to it. Stress commanders and protest
may be caused by just the mere presence of force, or by creating a greater group leaders are
demand than someone has resources to meet. Also, individual needs—food, increasingly seen as the
sleep, etc, can cause stress. Such stress can negatively affect an individual’s best means of preventing
motivation to pursue their goals. bad outcomes in crowd
s Focus – The ability to concentrate on those things leading to a goal. This situations. This is called the
includes picking up cues that indicate actions needed for goal accomplishment. negotiated management
This factor more immediately influences behavior than previous ones. It can model of crowd control… .
be altered by ineffective focus—focusing on those things that interfere with,
or are irrelevant to goal accomplishment. Causing individuals to focus on
personal needs for example may do this. Increased stress can also diminish focus.
s Emotions – They ultimately dictate individual behavior. They guide decisions,
influence reactions to situations, and determine how individuals feel. Emotions
vary among individuals in a crowd and change rapidly with the situation.
Physical or ego threats can negatively influence emotions. A physical threat
might be the size and resources of an opposing force. An ego threat might be
a goal that seems too difficult to accomplish. These threats can cause
frustration, anger, or despair.

Encouraging relationships and ownership
“It works when you can talk to them,” stated Dr. Waddington. Working relationships
between tactical commanders and protest group leaders are increasingly seen as
the best means of preventing bad outcomes in crowd situations. This is called the
negotiated management model of crowd control,16 and it is practiced by law
enforcement agencies in Washington, D.C., New York City, San Francisco, Los
Angeles, London, Paris, Zurich and Rome. It was also used in Somalia by U.S.
military leaders who repeatedly met with Somali warlords to resolve issues. As

16 McPhail, Schweingruber and McCarthy, 1998, op. cit.; John McCarthy and Clark McPhail. 1998.
“The Institutionalization of Protest in the United States.” Pp. 83-110 in David Meyer and Sidney
Tarrow (Eds) The Social Movement Society: Contentious Politics for a New Century: New York
and Oxford: Rowman and Littlefield; John McCarthy, Clark McPhail and John Crist. 1999. “The
diffusion and adoption of public order management systems.” Pp. 71-94 in Donatella della Porta, Crowd Behavior, Crowd Control
Hanspeter Kreisi and Dieter Rucht (Eds) Social Movements in a Globalizing World. London:
MacMillan Press, Ltd. 27

General Anthony Zinni USMC stated, “Always consider negotiations as a great
alternative to violence.”17

TALK
Working relationships between tactical commanders and protest group leaders
provide an opportunity to communicate. They allow group leaders to tell authorities
and tactical commanders what they want to accomplish. Additionally, it allows
authorities and tactical commanders to tell group leaders what they are prepared
to do, and how they might respond to cer tain crowd behaviors. Such
communications can do much to resolve issues and prevent violence.

TALKING TO SOMEONE IS BETTER THAN TALKING TO NO ONE
In some cases, tactical commanders may not be able to talk to all leaders before
a crowd assembles. Some groups do not have recognized leaders and are basically
ad hoc organizations. Other groups may have several leaders, but only some of
those leaders will negotiate. In such cases, meetings with some leaders may tell
you who the other leaders are and provide other critical information. Negotiations
also may encourage more moderate leaders to do things that will support the The police will encourage
tactical commander. demonstrators to protest
in an acceptable way and
BE THE INITIATOR offer favors to get them to
Many law enforcement agencies initiate communication with group leaders. Captain do just that.
Odenthal stated that the Los Angeles Sheriffs Department regularly attempts to
make contact with group leaders and work out issues well in advance of an event. Dr.
Waddington pointed out that the London Metropolitan police do the same and make a
concerted effort to maintain a working relationship with the leaders of groups, such as
labor and trade associations that hold demonstrations and protest marches.

WIN OVER THE DEMONSTRATORS
Communication also means persuasion. The London Metropolitan Police make a
concerted effort to “win over” demonstration leaders. They will often say, “We want
to help you to do this safely.” The police will encourage demonstrators to protest in
an acceptable way and offer favors to get them to do just that. Dr. Waddington
related an example involving some 100,000 striking coal miners in a protest
demonstrations that occurred in London’s Hyde Park. The police arranged parking
for the miner’s buses near the Park and transported labor leaders to Parliament.
By doing that, the police kept the demonstrators away from Parliament.

FOLLOW THE RULES
Several rules apply in negotiations. Credibility is key. Only communicate necessary
information and those actions which authorities and tactical commanders intend
to do. Negotiations should also be made from a position of strength. Through
negotiation an agreement might be reached. However, contingencies should be
developed in the event the agreement is violated.

GET THEM TO POLICE THEMSELVES
Working relationships between tactical commanders and protest group leaders
17 Kenneth Allard, Somalia Operations: Lessons Learned, (Washington DC: NDU Press, 1995), p.
71 28

often result in protest groups policing themselves. This is one of the basic premises
of negotiated management. Police know that rank-and-file protesters are more
likely to listen to their own leaders, as opposed to listening to the police. Protest
organizers are likely to buy into a cooperative effort and agree to ground rules set
during pre-protest negotiations. Many seek to stay on the right side of the law.
Captain Odenthal provided an example. During several demonstrations, leaders
approached him to point out troublemakers so that LASD deputies could remove
them. Dr. Waddington provided another example. During a labor strike in London,
several strikers dispersed anarchists who attempted to “hijack” the demonstration.

CONSIDER POINTS OF INTERVENTIONS WITH CROWDS
Negotiations require expertise. Captain Odenthal related a case in which he used
expertise to diminish protest fervor. During negotiations, group leaders stated that
they would march five miles to demonstrate their cause. Captain Odenthal, however,
invited them to march seven miles, which they opted to do. This distance increased
fatigue and reduced motivation. Crowd behavior can be influenced in other ways.

To provide a framework for tactical commanders in negotiations, Dr. Taylor provided
another model for crowd behavior. Similar to the individual model, this crowd model
is also a pyramid, comprised of five group factors, which follow a logical and
purposeful progression.

Figure 4.
Potential points of intervention for the control of crowds.

Each factor represents a potential point of intervention in dealing with groups or
the crowd. However, this model needs to be used with caution because, as empirical


29

data have indicated, crowds are not homogeneous entities, crowds are not made
up of isolated individuals, but of “companion clusters”, which arrive, remain and
leave together, and crowd participants are not unanimous in motives. Nevertheless,
these factors can be help negotiations and are addressed in ascending order on
the pyramid.
s Identity – The foundation for a group’s intentions and behavior. They may be
soccer hooligans, strikers, or ethnic factions. Identity reflects their values and
their purpose in their existence. Once you understand who they are, you can
understand what they may do, as Dr. Waddington stated. Crowd identity is
self-renewing, though. It requires a continued sense of purpose and action.
Making these difficult to achieve can alter identity. Helping groups achieve
goals also can validate it.
s Leadership – Provides direction, offers structure and process, and rewards
efforts. The tactical commander should assess leadership for its effectiveness
and span of control. Leadership can be negatively and positively influenced.
It can be undermined by creating disorganization within a group or uncertainty
with respect to accomplishment of goals. It can be positively influenced by
convincing leaders to work with authorities, and by encouraging other factions
to do the same.
s Cohesiveness – Elements that bring and hold a crowd together. This might be
an overarching value. Leadership and communication can also enable
cohesiveness. Additionally, cohesiveness requires participants to be
cooperative, loyal and compassionate. Tactical commanders can disrupt
cohesiveness by removing leadership or disrupting communication. They can
also take advantage of cohesiveness by winning over leadership and ultimately
its followers.
s Communication – The acquisition and dissemination of information. It is needed
to provide clear vision and goals, as well as, convey decisions to the crowd.
Threats, stress and internal conflicts can disrupt communication. When people
are more concerned about self-preservation, they may be less focused on
receiving and sending information. Tactical commanders can also take
advantage of group communication by getting leaders to convey cooperative,
rather than confrontational information to the crowd.
s Decisions – These impact the directions, goals and outcomes of those in the
crowd. It involves evaluating options and selecting the best one that will achieve
a goal. Decision-making within a large group is difficult. It can be even more
so if time is limited, communications are disrupted, and certain stimuli create
negative emotional responses. Conversely, tactical commanders can positively
influence decision-making by meeting with leaders and convincing them to
do it their way.

Acceptability—Redefining What It Means to Win
U.S. forces will find themselves conducting crowd control in difficult situations.
They will be in a foreign country in which some local groups may not accept them.
They may have to impose order in an area, and protect residents. Moreover, they
will be under intense media and political scrutiny. As the practitioners on the Panel
pointed out, it is an environment not unlike that in which the police operate.

30

SECURE THE MORAL HIGH GROUND
Winning in this environment is not like winning in combat. U.S. forces can appear
to be invincible and formidable, but they also risk being portrayed as oppressors.
If so, they can loose by appearing to win. Groups that perceive themselves to be
oppressed will readily seek victimhood in an effort to gain the support of public
opinion. Winning in this environment is about seizing and holding the moral high
Winning in this
ground. As Dr. Waddington stated, “… you must maintain the authority and
environment is not like
legitimacy of what you are doing… .”
winning in combat.

REACH OUT
Projecting a favorable image will require outreach to local leaders and citizens. Dr.
Waddington stated that the London Metropolitan Police often invite local leaders
to observe briefings on upcoming public order operations in their community.
Outreach will also require developing relationships with the media. Captain
Odenthal stated that the Los Angeles Sheriffs Department invites the media to
accompany tactical commanders during crowd control situations. Such outreach Outreach… requires
gives the impression that the authorities have nothing to hide. Additionally, it developing relationships
provides an opportunity for opinion leaders to see the tactical commanders’ side with the media.
of an event.

REMEMBER THE “CNN EFFECT”
In this environment, tactical commanders must consider how actions will play
among several publics—local, allied, U.S. and international. The CNN effect is
inescapable. In some cases, it will be difficult to accommodate all of these publics.
The most important will be the U.S. public, which is key to continued support of
U.S. forces. When responding, tactical commanders must also ensure that the
The CNN effect is
“bad guys are really bad,” as Dr. Waddington put it. He cited the example of the
inescapable.
recent May Day Riot in Great Britain. Demonstrators desecrated Churchill’s statute,
and the Cenotaph memorial to the British war dead, which turned the public against
the demonstrators.

ENSURE WEAPONS ACCEPTABILITY
In crowd situations, the use of weapons and particularly non-lethal ones will be
key. Their use is not merely a technical issue. It is a political issue as well. The
same weapon that may be publicly acceptable in one situation, may be …the use of weapons and
unacceptable in another. particularly non-lethal
ones … is not merely a
Weapons acceptability depends on whom they are being used against. For example, technical issue. It is a
strong policing tactics such as dogs and riot batons have been used against soccer political issue as well.
hooligans in Europe without public outcry. Use of the same measures against
striking workers, or welfare mothers can and has evoked a strong outcry.

Acceptability also depends on when and where a weapon is used. In the U.S., law
enforcement agencies readily use blunt impact weapons. However, the British use
of blunt impact weapons has been condemned by the UN, Amnesty International
and Human Rights Watch Helsinki.


31

Additionally, acceptability depends on whose behalf the weapon is used. During
Apartheid, the South African Police used the same tactics and instruments as
other nations. However, they were used in support of an authoritarian regime. As
a result, their weapons took on a different complexion.

One in particular was the “sjambok”—a flexible riot baton. It never killed anyone. It
just gave a stinging welt. The sjambok, though, was associated with slavery. It was
similar to the whip, the preferred instrument of the slave master. After Apartheid,
the sjambok was banned, even though the South African Police still carried shot
guns and Uzi machineguns. Tactical units should follow
a non-lethal weapons
FOLLOW THE NON-LETHAL WEAPONS HIERARCHY hierarchy,
Acceptability of weapons will also depend on appropriate and proportionate
responses. Tactical units should follow a non-lethal weapons hierarchy, whenever
possible. This means that they should initially use non-lethal measures for passive
defense, to hold crowds at a distance. If violent behavior occurs, tactical units may
increasingly resort to more active and greater levels of non-lethal force.

This force should also appear to be “self-triggering.” In other words, the individuals
who are the opponent should appear to be initiating their own consequences. This will
be especially important in following the non-lethal weapons hierarchy. As the level of
force increases, so will the likelihood of injuries and negative public reactions.

This non-lethal weapons hierarchy might occur as follows. Barriers might be used
initially to maintain distance – almost like a castle moat. Such barriers might include
anti-traction polymers. If behavior turns violent, tactical units might use accurate
blunt impact weapons against selected targets. These might be non-lethal weapons
A crowd is likely to be better
that fire baton rounds. In the event a crowd attempts to rush an area, which is
controlled by means of
regarded as a point of no compromise or a “die in a ditch” area, non-lethal weapons
containment—confining its
that disperse a large number of blunt impact projectiles might be used. Such
activities to a given area.
weapons might include the Modular Crowd Control Munition, which is a non-lethal,
Claymore mine-like device.

CONTAIN RATHER THEN DISPERSE
Current crowd control orthodoxy often places too much emphasis on crowd
dispersal. However, forced dispersal could result in a crowd breaking up into a
multitude of groups that would scatter over a vast area. This could pose even
greater public order problems. A crowd is likely to be better controlled by means of
containment—confining its activities to a given area. Additionally, a crowd has a
limited duration. Its numbers are likely to diminish as individual needs take
precedence over those of the crowd.

Conclusion
The peacekeeping actions in Somalia, Haiti, Bosnia and Kosovo have provided
increased demand for non-lethal weapons and greater need for understanding
the dynamics of crowds. These experiences have caused the U.S. military to
reconsider their use both technically and tactically. Because the approaches and

32

objectives of crowd control are vastly different from those held for traditional combat
missions, the military must gain additional insight into the individual, social, and
cultural impact of crowd behavior. Peacekeeping forces must adopt a new model
in understanding, evaluating, and responding to crowds. It is the Panel’s belief
that the decision-making guideline for crowd control offered in this report is a
constructive first step in that direction.


33

Objective 3
DEVELOPMENT OF EDUCATION AND TRAINING
GUIDELINES FOR CROWD CONTROL

General
Effective decision-making in crowd control depends on educating and training
mid- to senior-level officers. This education and training could be based on case
studies. Case studies could be used to exercise prevention planning and crisis
decision-making. An organization’s ability to learn partially depends on knowledge
Using case studies…
management – capturing lessons learned and best practices. Using case studies
contributes to education
not only contributes to education and training, but also provides feedback on tactics,
and training, …provides
techniques, and procedures. Ultimately, such a system can help an organization
feedback on tactics,
respond effectively to the challenges of crowd control.
techniques, and procedures.

Discussion
The Panel considered the education and training needed to support effective
decision-making in crowd control. The Panel believed that a broader program is
needed than currently exists. It should be oriented toward mid- to senior-level
officers in operational units that might be tasked with executing crowd control
operations. The Panel then considered the best means of education and training
in crowd control. This was seen as the case study. Additionally, a knowledge
management system is needed to build these case studies as well as contribute
to organizational change with respect to crowd control.

The Need for Case Studies
The case study is an excellent means of education and training in crowd control.
Sufficient information does not exist to develop models with validated crowd
behaviors and motivations. Additionally, artificial scenarios do not have any real
meaning to operators, as Major Ijames stated. Case studies with hard evidence
from real events do.

The case study is the basis for existing crowd control education and training
programs. The Los Angeles Sheriffs Department and other law enforcement
agencies around the county use case studies to prepare decision-makers for crowd
control situations. The military has also used them. Lieutenant Heal stated that
Marine Corps officers viewed videotapes of crowds to prepare for the evacuation
of UN forces from Somalia. They had to indicate the agitators, when they would
intervene, and what might happen next.

Such case studies could consist of several elements. They might include written
accounts of crowd control situations. These might come from official records that


34

were kept on actual events. For example, some training in law enforcement is
based on the Los Angeles Sheriffs Department’s accounts of the first two hours of
the 1992 riots. Case studies might also include best and worst practices – what
tactical commanders have found to work and not work in given situations. These
might come from after action reports.

These case studies should also include videotape footage of crowd control events.
As Major Ijames stated, plenty of videotape footage exists on public order operations
around the world. Dr. McPhail, who has produced and analyzed extensive film and
video records of behavior in crowds, also pointed out the value of this footage:
“There’s nothing more useful than to be able to see it and hear it.”

How to Use Case Studies
Crowd events should be broken down into elements. The Los Angeles Sheriffs
Department does this in its training. The examination of the initial elements of a
crowd control situation should be used to support prevention planning. Personnel Decision-makers should
should indicate what might go wrong in given situations, and how it might be know the employment of
avoided. This assessment should consider the intelligence, operations and logistics non-lethal weapons and
needed to avoid undesirable outcomes. The case studies could also be their effects.
accompanied by and evaluated within the framework of the decision-making
guideline proposed in Objective 2 of this report.

Case studies should be used to support crisis decision-making. Captain Odenthal
stated that in the training conducted by the Los Angeles Sheriffs Department,
personnel are given a situation and a requirement to respond within a stated time
period. As the course continues, personnel are given other situations and shorter
times to respond. According to Captain Odenthal, these time compressions are
the one factor that can be altered in case studies. They are intended to stress
personnel and accelerate their thought processes.

These case studies should also be “red-teamed.” Personnel should be designated
as protestors or crowd leaders. This concept was initially recommended by Dr.
Taylor and later endorsed by other Panel members. The use of opposing role Case studies should
players helps broaden perspectives and increases understanding of the oppositions emphasize respect for
thought processes, as Dr. Waddington points out. They may be used in negotiation human rights in crowd
and crisis response training. control situations.

Case studies must also be coordinated with skills training. Decision-makers should
understand the tactical skills being taught to troops that will operate in the streets.
Decision-makers should know the employment of non-lethal weapons and their
effects. Additionally, those troops employing non-lethal weapons should understand
the context and objectives of their employment.

Case studies should emphasize respect for human rights in crowd control situations.
The Panel unanimously underscored this. Dr. Waddington stated, “you are not
dealing with your people,” when forces deploy overseas and some service personnel
may believe they have greater license than they do. For example, he stated that it

35

had been alleged that British soldiers in Northern Ireland had sometimes put razor
blades in the rubber baton rounds, because the “Paddies” were not their people.

Respect for human rights is also a problem in other countries. Captain Odenthal
stated that a fellow U.S. law enforcement officer returned early from Bosnia after
trying to establish a police force there because he found it difficult for local police
to grasp the human rights concept. As Dr. Waddington stated, “The UN Declaration
is your bottom line. This is what you have got to live up to.”

Additionally, these case studies, and education and training in general should be given
to personnel that will use it. While this seems obvious, Panel members stated that the
wrong personnel are frequently trained in law enforcement organizations, and the
same may occur in the military. As Major Ijames stated, “You can do the training in the
war colleges, but that doesn’t mean that those officers will be the ones using it in
Kosovo.” Some law enforcement agencies, therefore, require tactical commanders to
have specific education and training before executing public order operations.

Knowledge Management—Lessons Learned
Education and training in crowd control will depend on a knowledge management
system. This is an organization’s ability to capture lessons-learned, and best and
…lessons learned should
worst practices. This knowledge management system is needed to build case
address failures as well
studies needed for education and training.
as successes.
The problem with crowd control knowledge is that it is limited. The military has
only recent experience in crowd control. In law enforcement, crowd control expertise
resides with relatively few individuals. When this expertise is translated into practice
in a crowd control situation, the lessons may or may not be adequately recorded.
As a result, knowledge of crowd control is not always shared easily and it fades as
individuals depart the organization.

Lessons learned, or after-action reports, are needed to build this knowledge
management system. U.S. forces prepare such reports after a major incident or
operation. However, capturing crowd control lessons will require established
reporting criteria. Major Ijames stated that in the Springfield, Missouri Police
Department, commanders must file after action reports on such crowd events as
the city’s Fourth of July fireworks display, in which 75,000 people attend. Their
report compares the pre-incident plan with the occurrence. It also provides the
pluses and minuses of the response, and recommends future actions.

These lessons learned should address failures as well as successes. The tendency
in many organizations is to scrutinize failure, but not look too closely at success.
Some successful outcomes, however, have all the elements of failure, but go well
for some unnoticed reason. This can create misperceptions about what actually
works in crowd control situations.

Lessons learned should be honestly reported. As Dr. Waddington stated, “You’ve
got to ensure in debriefing that people are honest and that they have the security

36

to be honest. They can’t think that if you screw up, your career is on the line.” Personnel
must be able to objectively evaluate the effectiveness of a tactical response, and
dispassionately determine why and how the particular outcome resulted.

Lessons learned should include video recordings. There is often a difference
between what people perceive and what actually occurs. By virtue of where
observers are located, they can only see a limited segment of all that is taking
place throughout the entire gathering. Video recordings have the same limitations.
Only if you have multiple video cameras distributed above, around and throughout
the gathering in question can you begin to get something close to representative
portrayal of the event. However, any video record is better than none because it
can be viewed, studied and analyzed repeatedly.

A knowledge management system can have other benefits in addition to
contributing to the development of case studies. It can help the organization act
faster and more appropriately. It can provide a source of information that
organizational members can quickly draw on in crises, so they do not always have
to “reinvent the wheel.” Additionally, a knowledge management system can provide
feedback on crowd control tactics, techniques, and procedures. It can help an
organization reject, validate, or improve these practices. Ultimately, a knowledge
management system can enable an organization to more effectively respond to
the demands of crowd control.

Conclusion
The military and law enforcement agencies must capture their experiences in crowd
control situations. Until more research can be conducted on crowd behavior, these
experiences provide the only reliable means of educating and training personnel
in crowd control decision-making. Capturing these experiences will depend on the
development of a knowledge management system, which establishes reporting
criteria. Such a system will help organizations to better understand, plan, and
execute crowd control.


37

Objective 4
DEVELOPMENT OF A “ROAD MAP” FOR FUTURE
RESEARCH ON CROWD BEHAVIOR

General
No substitute exists for research. Relative to personal expertise, research provides
the most reliable information on crowd behavior. More research is needed to
address the unknowns of crowd behavior. As new and innovative theories emerge,
there is an opportunity to study traditional approaches versus these new theories
to advance the knowledge base. Such research should be designed to compare
and contrast conflicting theories of why crowds turn violent. Ideally, sufficient
information would allow crowd behavior to be modeled with fairly accurate and
repeatable probabilities. To get to that point, the Panel recommended that research
be conducted to answer the following questions:
s Why people assemble?
s What social organizational structures exist within a gathering?
s What determines the intensity or “energy” of a gathering?
s How do crowds in other cultures behave?
s What psychological and social factors contribute to the emergence of
violent behavior in crowds?
s What situational and cultural factors contribute to the emergence of violent
behavior in crowds?
s What military or law enforcement actions lead to violent behavior in crowds?
s What crowd-control techniques are most effective in preventing and
stopping violent behavior?

Discussion
The Panel assessed the current knowledge regarding crowd behavior. It then
considered a possible long-term goal, the modeling of crowd behavior. Research
was seen as critical to getting to that point. The Panel then determined those
issues which future research should address to reach this end.

The Need for Research
Too often in the past, beliefs regarding crowd behavior were based more on
perceptions than research. Moreover, there is often a difference between what
individuals think they saw and what actually occurred. Dr. Waddington underscored
this point with his observations of the Poll Tax Riot. He stated that he was in the
midst of this riot and feared what he perceived to be violent mass behavior. However,
he later viewed videotapes of the riot, and noticed that only about ten individuals
in a section of the gathering were really engaged in violent behavior. As he pointed
out, data are more reliable than perceptions.


38

Research addresses the specifics of crowd behavior. As Dr. Waddington further
pointed out, most people tend to generalize about subjects, until they are asked to
detail certain cases. The collection of such details can help social scientists draw
more accurate conclusions regarding crowd behavior.

Additionally, research means perpetuity. Research often confirms what law
enforcement experts intuitively know about crowd control. This personal knowledge
is highly perishable, though. It degrades when individuals do not continually work
in this area, and it can be lost when individuals leave an organization. Research
findings last and they can be updated. They also can be more readily shared than Research often confirms
personal experience. what law enforcement
experts intuitively know
A “Road Map” for Future Research about crowd control.
The question is what research must be done to get to that point? The Panel
examined this question and produced a “road map” for future research on crowd
behavior. This road map recommends research be conducted on the following
subjects:

WHY PEOPLE ASSEMBLE?
Information has been collected on where and when people gather. However, why
they gather still remains to be determined. Dr. McPhail examined the question
using a secondary statistical analysis of 12 studies on riots. However, the issue
has not been examined first hand. The question of why people assemble is The question of why people
important because it indicates the participants’ level of commitment. Researching assemble is important
this task may be fairly simple. It would entail asking individuals at protests or because it indicates the
demonstrations “why are you here?” participants’ level of
commitment.
WHAT SOCIAL ORGANIZATIONAL STRUCTURES EXIST WITHIN
A GATHERING?
Research shows that a gathering consists of several clusters or smaller groups.
Each is usually comprised of individuals that know each other – friends, family,
co-workers. However, the overarching organization must be studied as well as
any other levels of social organization within a gathering.

WHAT DETERMINES THE INTENSITY OR “ENERGY” OF A GATHERING?
Some gatherings seem to have more intensity or “energy” than others. The Seattle
demonstrations against the WTO seemingly had more greater than the
demonstrations that occurred during the recent Democratic National Convention Little is known about crowd
in Los Angeles. In fact, Panel members said they overheard participants say that the behavior in non-western
Los Angeles demonstration lacked energy. This question is closely related to the two cultures.
previous questions. Possible correlates (and, in principle, possible sources) of this
‘intensity’ or ‘energy” might be individual commitment, organization and preparation.

HOW DO CROWDS IN OTHER CULTURES BEHAVE?
Almost all data on crowd behavior comes from research conducted in the United
States and Europe. Little is known about crowd behavior in non-western cultures.
In some ways, crowds in non-western countries may follow similar patterns as


39

those in the United States and Europe. They are likely to assemble, engage in
collective actions within a temporary gathering, and disperse. There is some
evidence that this is the case. However, crowds in other cultures are likely to have
different repertoires of collective action. Their collective facing, voicing, manipulating
and transiting will likely differ. They also may be given to greater ethnic antipathy.
Cultural anthropologists who are familiar with the political traditions and actions of
non-western cultures may provide such information.

Additionally, crowd researchers in other cultures may have developed theories of
crowd behavior that may be worth investigating. According to Dr. Waddington, an
Indian social scientist argues that some individuals are riot specialists. They operate
as part of ethnic, religious and cultural networks. According to this social scientist,
leaders within these networks can quickly mobilize these riot specialists to
participate in disorder. He further believes that these riot specialists were a factor
…research should consider
in the massive killing of Sikhs after the assassination of Mrs. Ghandi.
what situational factors…
contribute to violent
WHAT PSYCHOLOGICAL AND SOCIAL FACTORS CONTRIBUTE TO
behavior
THE EMERGENCE OF VIOLENT BEHAVIOR IN CROWDS?
Research should identify what psychological and social factors in a crowd make it
prone to violent behavior. By understanding these factors, it may be possible to
predict the likelihood of violence and intervene more effectively.

WHAT CULTURAL AND SITUATIONAL FACTORS CONTRIBUTE TO THE
EMERGENCE OF VIOLENT BEHAVIOR IN CROWDS?
Building on the more basic question about culture, research should examine if
some cultures (i.e., religious, political) are more prone to violent behavior and
why. These findings would provide the military with increased sensitivity to the
cues that predispose violence in foreign countries. Additionally, research should
consider what situational factors may contribute to violent behavior. Several issues
Research should be
that are worthy of investigation include crowd size, crowd density, physical space,
conducted that examines
resources, and size of force.
what specific techniques are
most effective in meeting
WHAT MILITARY OR LAW ENFORCEMENT ACTIONS LEAD TO
crowd-control objectives.
VIOLENT BEHAVIOR IN CROWDS?
The actions of military and law enforcement have often been blamed for violent
behavior that arises in a crowd. Research exploring what specific actions by forces
can lead to violence, would be invaluable for tactical commanders to understand
and act upon so that their forces act as deterrents rather than instigators of violence.

WHAT CROWD-CONTROL TECHNIQUES ARE MOST EFFECTIVE
IN PREVENTING AND STOPPING VIOLENT BEHAVIOR?
At the heart of the JNLWD’s mission is identification of crowd-control techniques
including non-lethal weapons that are most effective in preventing and stopping
violent crowd behavior. Research should be conducted that examines what specific
techniques are most effective in meeting crowd-control objectives.


40

Conclusion
Research on crowd behavior is needed to achieve the Joint Non-lethal Weapons
Program’s goal with respect to crowd control. Specifically, it seeks the “the means
to influence the behavior and activities of a potentially hostile crowd….” However,
not enough is known about crowd behavior to effectively do that. Research is
needed to address its unknowns. Only when more is known about crowd behavior
can accurate and repeatable probabilities be made regarding crowd control
outcomes.


41

Conclusion
Crowd control has been a driving factor in non-lethal weapons initiatives. The
crowd situations in Somalia provided the impetus for U.S. forces to adopt non-
lethal weapons. Since then, crowd situations in Haiti, Bosnia and Kosovo have …this report is intended to
increased the demand for non-lethal capabilities. Moreover, non-lethal weapons help U.S. forces respond
are being developed for crowd control. The problem, however, is that crowd behavior more quickly and effectively
has not been closely studied within the defense and law enforcement communities. to crowd-control situations,
particularly in peacekeeping
Therein lies the contribution of this Human Effects Advisory Panel. It assembled operations abroad.
recognized experts in crowd behavior to examine this issue as it relates to crowd
control. These are practitioners and academicians who not only have knowledge
of the subject, but also first-hand experience with it in Los Angeles, Seattle, London,
South Africa and elsewhere. Their views are also independent. They are not actively
involved in defense programs related to non-lethal weapons, and have no vested
interests in such areas.

This Panel has provided findings on crowd behavior that are based on scientific
evidence and reinforced by the experience of practitioners. It has found that some
military and law enforcement concepts for crowd control are not well founded.
Some concepts are based on stereotypes that social scientists have proven
incorrect. Additionally, much remains to be learned about crowd behavior before it
can be effectively influenced and controlled.

The Panel has gone a step further. Based on these findings, the Panel has sought
to facilitate a new approach to crowd control. It offers a decision-making guideline
for crowd control, which emphasizes prevention and negotiations, rather than
confrontation. This decision-making guideline is supported by the Panel’s
development of education and training guidelines for crowd control. It is oriented
toward mid- to senior-level military officers, and relies on the use of cases studies.
Additionally, the Panel has provided a “road map” for future research. This report
is intended to assist the Joint Non-lethal Weapons Program in developing non-
lethal weapons systems for crowd control.

The Panel has made these recommendations for the purpose of increasing
understanding of crowd behavior and crowd control. These insights are needed in
the development and employment of non-lethal weapons for crowd control.
Ultimately, this report is intended to help U.S. forces respond more quickly and
effectively to crowd-control situations, particularly in peacekeeping operations
abroad.


42

Crowd control report

More Related Content

Viewers also liked (20)

Similar to Crowd control report (20)

Recently uploaded (20)

Crowd control report