Non-violent
1/28
Non-Violent Conflict
H˚avard Mokleiv Nyg˚ard
Peace Research Institute Oslo
March 13, 2015
Non-violent
2/28
Outline
What is nonviolence
Main findings from the literature
Preference Falsification
Tactics & Outcome
Correlation or Causation?
Non-violent
What is nonviolence 3/28
Demonstrations Iran, 1979 Sit–in Karachi, 2013 Hunger strike
Non-violent
What is nonviolence 4/28
’For me, nonviolence was not a
moral principle but a strategy;
there is no moral goodness in
using an ineffective weapon’.
Nelson Mandela
Non-violent
What is nonviolence 5/28
Anti-regime protesters
Non-violent
What is nonviolence 6/28
Independence protesters
Non-violent
What is nonviolence 7/28
Anti-occupation protesters
Non-violent
Main findings from the literature 8/28
Chenoweth and Stephan identified:
323 campaigns from 1900 to 2010
106 mainly non-violent
214 mainly violent
Non-violent
Preference Falsification 9/28
’A single spark can start a prairie fire’
Often revolutions appear as out of the blue, but
in retrospect seem inevitable (Kuran 1991)
Kuran distinguishes between private and public
preferences in authoritarian regimes
Preference falsification can make unpopular
regimes appear as robust
This makes it much harder to overcome collective
action problems
External shock can provide information – in
combination with a mismatch between private
and public preferences this is likely to generate a
cascade of protests
Non-violent
Tactics & Outcome 10/28
Violent vs. Non–Violent Tactics & Outcome
Stephan & Chenoweth (2008; 2011) are the first studies to compare
non-violent and violent strategies across the entire world (1900-2006)
They find that non–violent movements are almost twice as likely to
succeed as their violent counterparts
Gamson (1990) study American challenge groups and find the opposite:
Groups employing force are more likely to succeed than those refraining
from violent tactics
Quite a few small N studies examine the relationship between tactics and
outcome (Ackerman & DuVall 2000; Ackerman & Kruegler 1994; Schock
2005; Skarp 1973, 2005; Wehr et al. 1994, Zunes 1994, Kurtz & Asher
1999)
Non-violent
Tactics & Outcome 11/28
Main Finding
Non-Violent movements are almost twice as likely to succeed as their
violent counterparts.
0
30
60
90
Anti−Occupation Regime Change Secession
Goal
Count
Campaign
Non−Violent
Violent
0.0
0.2
0.4
0.6
Anti−Occupation Regime Change Secession
Goal
Percentage.Successful
Campaign
Non−Violent
Violent
Non-violent
Tactics & Outcome 12/28
Why Civil Resistance Works
A campaign (both non-violent and violent) will succeed if:
1 The regime (as a whole) is defeated
2 The regime is, due to the costs inflicted by the campaign,
better off if they accommodate the group
3 Loyalty shifts take place within one or several of the regime’s
pillars of support
(These are not mutually exclusive)
Non-Violent movements are more likely to lead to accommodation
and loyalty shifts: directly and indirectly
Directly: Through the nature of non–violent tactics, reducing
accommodation risks and increasing the likelihood of loyalty shifts
Indirectly: Being better at mobilizing large number of people which
increases the costs inflicted on government and members who make
up the most important pillars of support
Non-violent
Tactics & Outcome 13/28
Tactics & Mobilization
Large campaigns are better at inflicting costs
directly on the government or at their main pillars
of support
Non-Violent campaigns are on average
four times larger than violent campaigns
20 if 25 largest campaigns were
non-violent
According to Chenoweth & Stephan (2011) this is
the main reason why non–violent campaigns tend
to be more successful than violent campaigns
But violent movements might achieve a bigger
bang for the buck
Non-violent
Tactics & Outcome 14/28
The Mobilization Advantage: Physical Costs
Many non–violent tactics demand no
skills other than sitting on the ground.
Non-violent
Tactics & Outcome 15/28
The Mobilization Advantage: Physical Costs
Many non–violent tactics demand no
skills other than sitting on the ground.
In contrast, participants in violent
campaigns, must be in good shape and
know how to use weapons. This take time
to learn, and for many is not possible
Non-violent
Tactics & Outcome 16/28
The Mobilization Advantage: Informational Difficulties
Individuals are more likely
to engage in protests
when they expect large
number of people to
participate
Information regarding
group size if easier to
access in non–violent
than violent campaigns
Violent campaigns –
particularly guerilla and
terrorist groups, have to
hide to survive
Figure: Naxalite guerillas trek back to their
jungle redoubt, south Bastar region
Non-violent
Tactics & Outcome 17/28
The Mobilization Advantage: Moral Barriers
Taking up weapons and killing adds a new dimension.
While some people might not have any moral difficulties taking up
weapons, most people have
People joining a non–violent movement often face no moral barriers
Non-violent
Tactics & Outcome 18/28
The Mobilization Advantage: Commitment Barrier & Risk
tolerance
People with varying degrees of risk and commitment tolerance can
participate in non–violent movements
It is possible to join a non–violent movement just for a day – this is
seldom an option in violent campaigns (training needs, access to classified
information)
It is easier to combine non–violent activities with a job and a family: Do
not have to make major life sacrifices (lower opportunity costs)
Non-violent
Tactics & Outcome 19/28
Why are Large Movements more likely to Succeed?
Large movements inflict higher costs on the government (This is true
both for non–violent and violent campaigns)
This makes it more costly to preserve status quo, and more attractive to
grant concessions
It also makes it more likely that some sort of loyalty shift take place
within the government’s most important pillars of support
Non-violent
Tactics & Outcome 20/28
Loyalty Shifts
Every political leader depends on some kind of
coalition (This is true for Kim Jong–un as well as
Erna Solberg)
The most important pillar of support is
the security apparatus (military and
police)
Without the support of the security
apparatus the days of the regime will be
counted
The economic elite as well often play an
important role:
Wood (2000) argues that the
anti–apartheid movement in South
Africa succeeded partly by inflicting cost
on the economic elite which responded
by pressing the regime to negotiate
Non-violent
Tactics & Outcome 21/28
Loyalty Shifts within the Security Apparatus
When demanded to use force to quell
the protests the security force
(military and police) might not always
do so. Instead they might:
Shirk: Do nothing and let the
protesters do what they want
(Extremely unlikely to happen
if the campaign is violent)
Defect: Join forces with the
campaign
Non-violent
Tactics & Outcome 22/28
Loyalty Shifts within the Security Apparatus
Repression is more likely to backfire if the campaign is non–violent
The military might be less willing to repress large number of people
With a large number of participants, the chances for kinship ties or other
social networks linking members of the military to members of the
movement increases. This is likely to make soldiers more reluctant to
shoot at protesters, particularly if they non–violent
More convincing that the military will be treated well by the new regime
if the campaign is non–violent
Non-violent
Tactics & Outcome 23/28
Summary
Non–Violent campaigns are more than twice as likely to succeed
Non–Violent campaigns better at mobilizing large number of people
The regime is more likely to negotiate with a non–violent than a violent
campaign
Non–violent campaigns are more likely to cause loyalty shifts within the
regime as well as amongst the regime’s erstwhile supporters
Non-violent
Correlation or Causation? 24/28
Does Correlation imply Causation?
Would violent campaigns be more successful if they made use of
non–violent strategies?
Figure: Anti–regime Protest against President Marcos in Philippines
Non-violent
Correlation or Causation? 25/28
Puzzling Tactical Choices?
Research by Cunningham, Bakke & Seymour (2012), Pearlman (2011) &
Bloom (2005) suggest that organizations engage in two contests
simultaneously:
1 Against the state, and
2 Against other organizations fighting for the same group
Violence is an efficient way to ensure that an organization remains
relevant
As competition between groups increases organizations are more likely to
shift to violent tactics. This might happen despite the group as a whole
(including all organizations) would be more likely to succeed against the
state if they sticked to non–violent tactics
This can explain why organizations make use of tactics that do not
increase the likelihood that the group as a whole will succeed
Non-violent
Correlation or Causation? 26/28
Is there a selection bias?
Would violent groups with a small mobilization potential
and/or are located in the periphery be better off using
non–violent means?
Far away from any centre non–violent tactics tend not to be
very efficient
Over time non–violent tactics tend not to be very mobile
In these situations violent tactics might be more efficient than
non–violent: This might explain the lack of non–violent
secessionist movement
Non-violent
Correlation or Causation? 27/28
Available Data
Non-Violent and Violent Data
NAVCO 1.1: Campaign oriented (1900–2006)
NAVCO 2.0: Campaign-year oriented (1945–2006)
NAVCO 3.0: (To be released) Event data (1987–2011)
MAROB: Ethnopolitical organizations in North Africa and the
Middle East (1980–2004)
Violent Data
UCDP/PRIO Armed Conflict Database: Conflict-year oriented
(1946–2012)
Correlates of War: Conflict-year oriented (1816–1997)
Non-violent
28/28
’Erich, we can’t beat up hundreds of thousands of people’.
Erich Mielke, Head of Stasi (1957 – 1989)
back