Ballots With Bullets - The Rise of The Extreme Right in Turkey

LANDS AND
bs_bs_banner
BALLOTS WITH BULLETS: ELECTIONS, VIOLENCE,

AND THE RISE OF THE EXTREME RIGHT IN TURKEY
Şefika Kumral
Dominant literature on social movements and extreme right politics assumes that participation in electoral
politics is incompatible with utilization of political violence. This elections-violence dualism has recently been
challenged by the rise of far right political movements and parties that utilize political violence and electoral
politics in various countries, such as the BJP in India, the Golden Dawn in Greece and the MHP in Turkey.
This paper examines the negative case of the extreme right in Turkey, led by the MHP (Nationalist Action
Party) to explain how far right movements can utilize political violence and avoid both marginalization and
state repression. The paper utilizes mixed methods using original quantitative data on extreme right
violence in Turkey compiled from historical newspaper archives and in-depth interviews conducted by the
author. Through a combination of historical and quantitative analysis, the paper shows that political violence
used by u €lk€
uc€
u movement has not declined but further increased with extreme right votes. Qualitative
analysis of in-depth interviews explains how extreme right was able to escape marginalization and state
repression due to the political context marked by Kurdish conflict as well as movement’s strategic utilization
and framing of political violence in response to this political context.
Introduction
The rise of extreme right parties and movements in liberal democracies in

recent decades has been recognized as one of the most alarming political devel-
opments. Once coined as “shadows over Europe,”1 extreme right parties man-
aged to get “durable electoral support” in many European countries, showing
that they are not ephemeral actors in European politics. In various countries,
including France, Netherlands, Switzerland, Denmark, Italy, Norway, Hungary,
Austria, and Bulgaria, extreme right parties have become significant challengers
of mainstream parties.2 The flood of European Parliament with Euroskeptic
parties after the recent elections has also been a clear indicator of the rise of the
extreme right. “Outright neo-Nazi parties” including the Golden Dawn of
Greece, Jobbik of Hungary and the German National Democratic Party even suc-
ceeded in gaining seats in the most recent European Parliament.3
Parallel to the electoral success of extreme right parties, extreme right vio-
lence has also escalated in the European socio-political scene, as indicated by
recent extreme right mobs against refugees, attacks in Norway and Odessa,
Journal of Labor and Society · 2471-4607 · Volume 20 · June 2017 · pp. 231–261
C 2017 Immanuel Ness and Wiley Periodicals, Inc.
V
232 JOURNAL OF LABOR AND SOCIETY
violent Islamophobic actions utilized by extreme right organizations such as

English Defense League in United Kingdom,4 and high levels of antiminority
and racist violence in various European countries.5 Rise of paramilitary violence
by “patriot groups” in the U.S. parallel to the rise of right-wing nationalism in
the electoral arena—as indicated by popular support of the Tea Party movement
and the presidential campaign of Donald Trump- shows that the United States
is not immune this globally alarming trend.
While both utilization of violence and electoral success is part of the resur-
gence of new extreme right movements, the general tendency in the mainstream
literature has been to analyze these developments almost as two separate trends
and to give primacy to the resurgence of extreme right political parties in the
electoral arena.6 This tendency is reinforced by the belief that utilization of vio-
lence declines as extreme right parties participate in electoral politics. In the
1980s and 1990s, this perception had a strong empirical basis in Europe where
electoral success of extreme right parties was highest where extreme right vio-
lence was low.7 These empirical findings strengthened the assumption that vio-
lence and electoral success were largely incompatible strategies of extreme right
political action. In this context, many scholars believed that “[w]hile political
extremism of all stripes may generate violence and hatred, it tends not to make
large electoral inroads. Skinheads do not win political campaigns.”8
A similar tendency of treating violent and nonviolent forms of collective
action largely as mutually exclusive and incompatible action repertoires (or some-
times as alternatives to each other) has been prominent in the contemporary
social movement literature. A number of studies in this broader literature main-
tain that social movements tend to radicalize and rely on violence when they
start to lose power and demobilize.9 Some scholars provide movement-level
explanations for this shift in strategy, suggesting that reliance on organized vio-
lence is largely adopted to “compensate for the loss of overt support.”10 It has
also been suggested that the negative association between use of violence and
popular support is mediated by state repression. According to many social move-
ment scholars, social and political movements that rely on nonviolent strategies
might switch to violent strategies if they face state repression.11 Other scholars
suggest a more dynamic relationship by arguing that use of violence by social
and political movements would also trigger state repression and would further
marginalize these movements.12
All these different explanations see some kind of dualism or incompatibility
between violent forms of action repertoires and nonviolent ones, including elec-
toral participation. These explanations, however, cannot account for the con-
temporary rise of far right parties and groups around the world who have
managed to gain major electoral successes despite their explicit use of political
violence on the streets either. After the 2008/2009 crisis, for instance, the Golden
Dawn, a fringe party in Greece with neo-fascist characteristics, managed to
make inroads into parliament despite its openly violent political record. In the
same period, the British National Party—which recruits members from skin-
head groups and promotes racist violence—secured a seat in the London
KUMRAL: BALLOTS WITH BULLETS 233
Assembly in 2008 and gained two seats in the European parliament in 2009
European elections. Outside Europe, India’s latest elections also witnessed the
historic success of the BJP, a right-wing Hindu-nationalist party which has been
involved in various incidents of communal violence against the Muslim commu-
nity.13 These cases whereby extreme right parties/movements have been gaining
electoral successes while sustaining use of violence, start to challenge perspec-
tives which perceive electoral popularization and use of violence as mutually
exclusive political action repertoires and raise an interesting puzzle: How do
extreme right wing groups maintain their popular support and avoid state
repression and further marginalization while utilizing violence?
This paper studies the rise of extreme-right Nationalist Action Party
(Milliyetçi Hareket Partisi, MHP) in Turkey since the 1980s, which utilized vio-
lence and electoral strategies as complementary form of politics. Although the
movement is considered to have entered a period of moderation since the 1990s,
quantitative and qualitative analysis presented in this paper demonstrate that the
movement actually increased its popular appeal while utilizing violence in an
increasing scale in the same period. This violence was directed to the leftist
youth, socialists, and communists in the 1990s, and increasingly to the proletar-
ianized and politicized Kurdish forced migrant population in the 2000s. Con-
trary to the expectations of the dominant scholarship on social movements and
far right politics, the u€lk€ u movement managed to avoid marginalization and
uc€
state repression, and further increased its political power in the electoral arena
while utilizing political violence. After demonstrating this trend, the paper dis-
cusses three mechanisms through which the u €lk€
uc€u movement avoids marginali-
zation and state repression: (1) discursive moderation, through which the
movement frames its actions and presents itself as less violent than it actually is
(2) interaction with the masses, whereby the movement pays specific attention not
to act as an isolated fringe/extremist group but to act in a larger crowd which
portrays violent incidents as reactions of larger masses, and (3) strategic alliance
with the state, whereby the movement does not directly engage in acts that would
destabilize the state and challenge state institutions and cooperate with police
forces when necessary.
Literature
The dominant literature on far right movements and social movement litera-
ture maintains that popular electoral support and violent action repertoires are
incompatible. As briefly summarized in the Introduction, the literature presents
three interrelated explanations for this incompatibility. The first is the modera-
tion thesis. If extremist movements participate in parliamentary elections and get
popular support, they will become more moderate parties and gradually stop
using violence. The second is the marginalization thesis: If far right movements
use violence, they will become marginalized and lose their popular support; or—
as some scholars claim—movements will rely on political violence when they
start to marginalize and to lose popular support. The third explanation is state
repression thesis, which builds on and extends the marginalization thesis by show-
ing the role played by state repression: If movements use violence, state actors
will repress them; and this will lead to the marginalization of the movement.
Using a combination of these explanations, the dominant literature predicts a
negative association between electoral participation and use of political violence.
Simultaneous rise of far right electoral popularization and violence around
the world in recent decades, however, has produced a set of alternative explana-
tions based on research from a number of negative case studies. A first set of
explanations focuses on the impact of socio-economic crises in the rise of fascism
and right-wing movements.14 This explanation is widespread in accounts of the
recent European experience, and focuses on the effects of economic crisis,
austerity-politics, poverty, and rising youth unemployment in the production of
anti-immigrant/xenophobic sentiments as well as alienation of the general elec-
torate from mainstream parties. These processes create a fertile environment for
fascist/neo-fascist parties to gain electoral successes and to use violence against
immigrants more easily thanks to the growing anti-immigrant/xenophobic
sentiments in the society. Hence, electoral popularization and violence are two
distinct consequences of the same socio-political and socio-economic transfor-
mations. These perspectives resonate with Karl Polanyi’s explanation of the rise
of fascism during the interwar period, socioeconomic deprivation theories of
radical right violence15 as well as ethnic competition theories.16 While these
perspectives explain the important link between the crisis of capitalism and rise
of far right movements, they fail to explain electoral popularization and violent
mobilization of far right parties in regions which were not heavily affected by
crisis/austerity politics or characterized by declining wages due to increases in
immigration. This problem becomes a critical one especially when examining
the rise of the far right in new zones of material expansion of trade and produc-
tion in the global South such as in India or Turkey.
A second set of explanations focuses on how violence against ethnic minor-
ities can be used by far right elites to reshape and activate ethnic boundaries
before elections in order to increase votes.17 Mostly derived from the experience
of the BJP/RSS in India—but also extended to cases such as Ireland, Malaysia,
and Romania—these explanations establish a causal relationship between the
rise of ethnic violence and electoral success of far right parties.18 Accordingly,
ethnic riots or other forms of far right violence are used as a deliberate political
strategy of electoral competition, designed by the elites to activate ethnic divi-
sions/boundaries that will move people to vote in line with their ethnic preferen-
ces, hence help far right elites increase their votes in the upcoming elections.
While filling an important gap in the literature by showing why political elites
utilize violence as a strategy, these perspectives does not necessarily explain how
certain movements can achieve electoral popularization via political violence
while others cannot. Furthermore, they do not pay attention to the reverse rela-
tionship: whether or not increasing electoral strength and popularization can
also be used to justify and normalize violence by far right parties.
The third set of explanations focus on the ideology, political agenda, and
organizational structure of extreme right and fascist parties. In classical fascism,
as illustrated by the historical Italian and German cases, electoral participation
and violence were not seen as two alternative sets of action repertoires, but as
complementary forms of politics to be used in advancing and protecting a nation
from its internal and external enemies. As Linz puts is, fascist parties “relied on
activist cadres ready for violent action combined with electoral participation to
gain power with totalitarian goals by a combination of legal and violent
tactics.”19 Likewise, Michael Mann emphasizes how paramilitarism and elec-
toral struggle were interlinked by drawing attention to “bottom-up” and popular
nature of fascist paramilitarism.20 These findings about historical fascist parties
have resonance today since many far right parties also follow in the footsteps of
these classical fascist parties in terms of their ideologies, political agenda, and
organizational structure. In the case of the rise of Golden Dawn, for instance,
Koronaiou et al. show that Golden Dawn’s young voters and supporters cannot
be reduced to angry youth with “an emotional reaction to the [economic] crisis”
but they have “wider ideological and political affinities and links [with fascism]
that have been building over the previous two or three decades.”21
While these perspectives point out that electoral participation and the use of
violence have never been mutually exclusive strategies for fascist parties, they do
not provide and explanation of the changing strategies and tactics of these par-
ties over time. More specifically, when fascist parties fail to gain control over the
state and face state repression, they often embrace a strategic discourse of mod-
eration to avoid repression. Then, it becomes very difficult for them to openly
use violence. How, then, do they manage to reconcile a moderate image/dis-
course and continue to utilize political violence?
The Negative Case of Turkish Far Right
In order to bring some light to these questions, this article brings another
negative case into the existing debates in the literature by examining the u €lk€
uc€
u
(idealist) movement in Turkey—led by Milliyetçi Hareket Partisi (Nationalist
Action Party, MHP)—from 1981 to 2013, with a focus on the relationship
between electoral participation and utilization of violence. In comparative-
historical studies, negative case methodology emphasizes “relation between
theory and history” and aims to broaden the scope of explanatory power of a
theory by studying cases where a particular outcome predicted by existing theo-
ries did not occur.22 Put differently, this method posits that anomalous cases
that are not consistent with the theory provide the researcher with information
that would lead to the expansion of existing explanations.
The negative case of Turkish far right advances our understanding of the
relationship between electoral popularization of far right movements and their
use of violence in three different ways: (1) Criticizing elections-violence dualism
in the literature, it presents an understudied negative case whereby electoral
popularization of a far right party does not necessarily result in a decline in use
of violence; but violence and electoral popularization can reinforce each other
under certain conditions. (2) It illustrates how the dual rise of the far right votes
and violence can take place in a new zone of material expansion of trade and pro-
duction, where socioeconomic misfortunates or crisis do not have much explan-
atory power. (3) It also exemplifies how an ultranationalist far right party can
change strategy on state repression, start using a moderation discourse to avoid
repression once again, and manage to reconcile this new moderate image with
its use of violence.
Officially founded in 1969, the MHP is a Turkish ultranationalist party,
which has been compared to classical fascist parties in terms of its political pro-
gram and strategy,23 its organizational structure,24 and its ability to openly lead
“paramilitary organizations fighting on the streets.”25 Like fascist parties of
interwar Europe, which established and maintained close ties with civil society
organizations in their rise to power, the MHP also established and preserved
informal but very tight organic ties with various civil society organizations. The
most well-known, widespread, and important of these were youth organizations
€ u Ocakları (Hearth of Idealists).26 Ulk€
called Ulk€ € u Ocakları activists were not for-
mally linked with the party but openly embraced the party’s ideology, program
and leadership. In the 1970s, the u €lk€ u activists were primarily known for their
uc€
nationalist propaganda among the youth and their infamous association with
political violence against the leftist movement.27 In late 1970s, on average 22
people died a day because of political violence, the majority in clashes between
the u€lk€
uc€u militants and leftists.28 This intensification of political violence and
increasing instabilities paved the way for the military coup on September 12,
1980.
After the 1980 military coup, together with other political parties and organi-
zations, the MHP was banned and activities of u €lk€u ocakları were suppressed by
the military junta regime. With the gradual lift of these bans in the course of the
1980s, however, the u €lk€ u movement started to re-organize. From the 1990s
uc€
onwards it rose from its ashes, started to gain major successes in both national
and municipal elections, and became the third largest party in Turkey. Because
of the high levels of electoral support that the MHP started to receive beginning
in the 1990s (see Figure 1), this new chapter of extreme right politics in Turkey
has widely been seen as the exact opposite of the 1970s, with its high levels of
violence but low levels of electoral support.
Parallel to this post-coup electoral popularization, the MHP and the u €lk€
uc€u
movement embraced “discursive moderation.” In their public speeches and dec-
larations, both party and movement leaders put an emphasis on “refraining from
violence.” Contrary to what has widely been assumed, this discursive transfor-
mation did not start with the new leadership under Devlet Bahçeli, but started in
the early 1990s when the movement was still led by its historic leader, Alparslan
T€urkeş. In line with the party leadership, u €lk€
uc€u activists started to publicly
declare that they did not wish to turn back to the 1970s and embraced a dis-
course of moderation.
Figure 1. Votes for the MHP in National and Local Elections, 1969–2014.
Source: Turkish Institute of Statistics.
At first sight, the MHP’s repression by military junta regime after the 1980
coup as well as its “electoral turn” and “moderation” in the 1990s are in line with
the expectations of social movement scholars, who maintain that violent move-
ments switch to nonviolent strategies in response to state repression. The bulk
of the scholarship on the electoral popularization of the MHP also echoes this
explanation, emphasizing the transformation of this once extremist movement
to a centrist/moderate party in the post-coup period.29 While most of these
analyses are based on the transformation of MHP’s public discourse, almost no
scientific study so far has provided an analysis of violence associated with the
u
€lk€ u movement in this so-called “moderation” period.
uc€
A systematic analysis of news reports of u €lk€ u associated violence and in-
uc€
depth interviews conducted in districts with high levels of communal violence
reveals that electoral popularization of the MHP in the 1990s period was accom-
panied by increasing utilization of political violence by the movement. While
violent events associated with the u €lk€ u movement were very low during the
uc€
military junta regime in the early 1980s, these violent events rose dramatically in
the course of 1990s, temporarily declined between 1999 and 2002, and acceler-
ated rapidly after 2002 (see Figure 2).
Precisely for this reason, it would not be accurate to describe the historical
trajectory of the u €lk€ u movement as a story of moderation, moving from an
uc€
extremist-fringe movement heavily relying on political violence to a centrist,
popular force with major electoral successes as illustrated in Figure 3a. Rather,
recovering from the suppression of the 1980 coup d’etat, the u €lk€ u movement
uc€
managed to position itself as a movement that receives popular support and
gains major electoral successes while maintaining its use of violence in the
streets. This trajectory is illustrated in Figure 3b, which does not conceptualize
the relationship between elections and violence on a singular axis but on two
independent different axes.
This particular trajectory of the far right in Turkey from 1980 to the present
opens up important questions that are relevant for existing discussions in the lit-
erature: (1) What is the relationship between u €lk€ u-associated violence and
uc€
€ uc€
Figure 2. Electoral Strength of the MHP and Frequency of Violent Events Associated with the Ulk€ u
30
Movement.
electoral popularization of the MHP in the post-coup era? (2) Which factors
contribute to the dual rise of electoral power and violent mobilization of the
Turkish far right? (3) How does the u €lk€ u movement manage to avoid margin-
uc€
alization from society as well as state repression despite its use of violence?
I argue that in the Turkish case, electoral popularization of the MHP and
violence used by the u u ocakları have a mutually reinforcing relationship. What
€lk€
created this reinforcing relationship, however, is not the effect of the economic
crisis or austerity politics, but a unique political context shaped by the rising
Kurdish politics in Turkey. In the course of the recovery from the suppression
Figure 3. Historical Trajectory of Far Right in Turkey: Two Perspectives.

of the military coup, the u€lk€ u movement managed to use the escalation of the
uc€
Kurdish armed and democratic mobilization to propagate the idea that “internal
enemies” are getting stronger and the idea that both in the parliament and in the
streets the u
€lk€ u movement is protecting the motherland. To avoid state repres-
uc€
sion, both the MHP and the u €lk€ u movement embraced discursive moderation.
uc€
They publicly announced that they have changed and they do not want to turn
back to the civil-war like environment of the 1970s. Yet, when they used vio-
lence, they framed it in a way to make these events appear less violent than they
really were, offering justifications for the use of violence—for example, “internal
enemies (i.e., the PKK and its allies) are getting stronger”—that resonated with
potential electoral constituents as well as with the political concerns of the exist-
ing governments. The MHP’s growing electoral strength also helped the u €lk€
uc€u
movement present itself in the eyes of the larger masses not as a fringe/extremist
movement but as a “responsible” political movement doing what it takes to
protect the motherland. Furthermore, the u €lk€ u movement avoided state
uc€
repression by paying specific attention to acting within a larger crowd in these
violent episodes, not targeting state institutions, and avoiding situations that
would directly destabilize the state.
Data and Methods
This paper uses a combination of quantitative and qualitative analysis to

show the relationship between elections and violence in Turkey. Using mixed
methods has increasingly been used in studies of communal and ethnic violence
in order to be able to address different interrelated segments of the same
research problem.31 In contrast to conventional approaches in the literature,
however, in this paper, statistical analyses will be used to identify general pat-
terns—instead of causal relationships—between electoral popularization and use
of violence. After presenting these general patterns, the paper will use qualitative
analysis based on original interviews conducted by the author to provide with an
explanation to how this dual rise of electoral popularization and the use of vio-
lence is possible.32
Quantitative Analysis
In the existing literature on far right movements in Turkey, there is no exist-

ing dataset for political violence associated with right wing movements in gen-
eral or the u
€lk€ u movement in particular. For this reason, I compiled a dataset
uc€
of right-wing nationalist violence incidents in Turkey—the Ethnic and
Nationalist Violence in Turkey (ENViT) dataset—using historical archives of
Milliyet daily newspaper.33
Using news reports to analyze diverse forms of social unrest is a common
and respected strategy in the social movements literature34; literature on ethnic/
racial conflict and communal violence35 as well as literature on radical right-
wing and racist violence.36 For scholars studying social movements, sometimes,
newspapers are the only possible choice due to the lack of official or alternative
statistics on many types of social movements. Even in the existence of official sta-
tistics, however, various scholars have suggested that newspaper sources can be
more reliable than official statistic on measuring different forms of social
unrest.37 Likewise, in literature analyzing ethnic and communal conflicts news-
papers are often preferred over other methods of gathering quantitative data
such as household surveys.
Data collection strategies that rely on news reports, however, are not with-
out limitations. On the contrary, one must be very careful while analyzing social
movements through news reports, especially with respect to potential biases
such as selection and data collection biases.38 As long as the researchers are
aware of these possible biases, however, they can implement strategies to mini-
mize their effects. For instance, in a study that focuses on far right nationalist
movements, a tendency to associate most forms attacks against laborers or ethnic
minorities with extreme right political actors without much evidence (a tendency
of most of the left-wing newspapers) or a tendency of not reporting extreme-
right wing violent incidents or the political identities of the perpetrators of these
events (a tendency of extreme right-wing newspapers) may affect findings. While
it is not possible to completely eliminate these selection and description biases,
one can minimize their effects by selecting news sources whose biases are mini-
mal (or much less compared to existing alternatives) and consistent across time
and space.39
The ENViT dataset includes diverse forms of instances of right-wing nation-
alist and collective violence originating at the societal level in Turkey from 1980 to
2011. The emphasis on societal level in this definition indicates that I exclude vio-
lence perpetrated by the state/government toward minorities from the scope of
analysis and focus on violence perpetrated by nonstate actors. The database cov-
ers instances of collective violence associated with u €lk€ u political actors, and/or
uc€
nationalist mobs/vigilantes directed against vulnerable minorities including
Kurds, Alawites, Armenians, Greeks, and the Roma population. My definition
of collective violence includes a wide spectrum of violent actions including riots,
lynchings, raids, violent attacks and clashes, beatings, bombings, and arsons.
This definition includes (a) instances of violence perpetrated by a group towards
other individuals or groups; (b) clashes between two or more groups; or (c) vio-
lence perpetrated by a group towards buildings like homes and institutions asso-
ciated with particular groups like minorities or rival political groups. This
means that violent conflicts between individuals are excluded from the defini-
tion. The definition of violence captures actual physical attacks or destruction of
property but excludes instances that involve discursive or symbolic forms of vio-
lence (such as booing and cursing) as well as group violence that is directed
against objects (like graveyards, monuments).
Using the ENViT dataset, l analyzed the relationship between electoral
popularization of the u €lk€
uc€u movement and violence used by the movement
using a set of covariates and control variables derived from the existing litera-
ture. The dependent variable, that is, the number of right-wing nationalist and
collective violence incidents associated with the u€lk€ u movement, is filtered

uc€
from the ENViT database. I define a violence incident as “€
ulk€ u-associated” vio-
uc€
lence if it was conducted by right wing nationalist actors associated with the
u
€lk€ u movement or the MHP, or if it was conducted by groups that use right
uc€
wing nationalist symbols such as the “bozkurt (greywolf) sign”—the sign of the
u
€lk€ u movement—or slogans that refer to the u
uc€ €lk€
uc€u movement.
Independent Variables. According to the existing “elections-violence dualism”

literature, as the electoral strength of the MHP increases, the level of u €lk€ u-
uc€
associated violence should decrease; and as u €lk€ u-associated violence increases,
uc€
votes for the MHP should decrease. According to critics of this approach, how-
ever, this is not necessarily correct. Theories which emphasize that violence can
be used as a strategic tool to increase far right votes expect that electoral strength
of the MHP should increase as u €lk€ u-associated violence rises. Yet these theo-
uc€
ries do not necessarily expect violence to increase as electoral strength of the far
right parties increases.
To assess the validity of these competing explanations, I tested the impact of
electoral strength on the levels of movement violence. I used electoral strength
of the MHP as one of the independent variables. Electoral strength of the MHP
is calculated as the percentage of votes MHP gained in national and local elec-
tions. Percentages of votes are derived from Turkish Statistical Institute (http://
www. tuik.gov.tr). In addition, I included a dummy variable measuring whether
or not there is an election next year to check whether an upcoming election, and
hence political/electoral competition increases the likelihood of u €lk€ u-associ-
uc€
ated violence as expected in the literature.40
In order to test effects of socioeconomic deprivation and ethnic competition
theories, I used annual GDP per capita levels, annual changes in the unemploy-
ment level, and average wages in the manufacturing sector.41 In the literature,
while all three measures are widely used to assess the level of socioeconomic
deprivation; unemployment level, and annual wages of industrial laborers are
also used to measure the effects of ethnic competition. According to socio-
economic deprivation and ethnic competition theories, both the frequency of far
right nationalist violence and electoral strength of far right parties increase as
GDP per capita and wage levels decrease, and the unemployment level
increases.
To test the validity of the explanations which associate the dual rise of far
right votes and violence in Turkey with increased perception of the “Kurdish
threat” which open up opportunities for the Turkish far right to mobilize masses
against the “internal enemy,” I produced a composite variable called “Kurdish
mobilization” using two distinct variables. First variable is the number of battle
deaths during the armed conflict between the Turkish army and the PKK as an
indicator of the level of ethnic secessionist warfare. The second variable is the
electoral strength of the pro-Kurdish parties and groups as an indicator of the
electoral mobilization of the Kurdish population, which is another dimension of
the perceived “Kurdish threat.” Composite variable that I call “Kurdish
mobilization” captures these two distinct dimensions as well as the transforma-

tion of the perceived “Kurdish threat.”
Two additional control variables were used in the analysis. The first is a
dummy variable showing whether or not the extreme right MHP is in the gov-
ernment that year. Many studies have shown that the level of extra-institutional
protests tends to decrease when movements have their representatives or allies
in the government.42 The second—which I also used in models assessing the
electoral strength of the MHP—is a dummy variable looking at the effect of
military rule, hence, state repression. This variable has a value of 1 in years when
there is a military junta in power (hence no participatory elections) and a value
of 0 in years when there is no military rule.43
Methods. All variables, whether drawn from the ENViT data or from official
statistics, are aggregated by year to construct a time series data set with 31 years
for the 1981–2011 period. Since the dependent variable—that is, number of
right-wing nationalist and collective violence associated with the u €lk€ u move-
uc€
ment—is count data with non-negative integer values that show properties of
over-dispersion (l 5 7.09; r2 5 40.62; also see Figure 4 below), I used negative
binomial regression analysis to estimate the regression coefficients, which is a
preferred method for scholars working on count data of ethnic violence and
social movements.44
Because I aggregate data by year, the total number of observations in the
analysis is small (n 5 30).45 Small number of observations constitutes a limita-
tion that this study shares with other main comparative-historical analysis of
communal and ethnic violence in the literature.46 This limitation partly arises
from the decision to conduct time-series analysis instead of cross-sectional anal-
ysis. In order to capture the historical transformation of the relation between
electoral popularity and violence of the u €lk€ u movement, I chose to focus on
uc€
variation over time rather than choosing one year and look at variation over
€ uc€
Figure 4. Histogram of Ulk€ u-Associated Violence Incidents Per Year.
space (e.g., cities and regions). Conducting a cross-sectional time-series or

panel-data analysis, which would account for variation in time and space while
further increasing the number of observations, is not possible due to lack of data
on the city or regional level for most of the key independent variables. Likewise,
increasing number of observations through aggregation of the data by month or
quarters (instead of by years) is not viable due to the lack of monthly/quarterly
data for most independent variables. Moreover, I decide not to transform the
yearly data into a monthly or quarterly data by reproducing the yearly value at
monthly or quarterly levels because this would reduce the variability of these
variables and, hence, artificially reduce their explanatory power.
Thus, similar to key comparative-historical analysis of communal and ethnic
violence in the literature, I conducted a time-series analysis, even if it means hav-
ing a relatively limited number of years/observations. For the purposes of this
paper, however, this limitation is not very damaging because the statistical find-
ings are not presented as evidence of causal relationships. On the contrary, as
mentioned above, statistical findings are utilized to reveal general patterns and
statistical associations that point out existing gaps and discrepancies in the litera-
ture. The main explanation for these patterns will be provided through a histori-
cally grounded analysis of qualitative data. It must also be noted that the main
limitation of using small number of cases is that it increases Type II error, mean-
ing we may fail to find significant results even though there might actually be
significant effects. This limitation is not very damaging for our purposes because
the findings we will pay attention to are those which are still significant despite the
small sample size.47
Interview Data
Although the statistical analysis presented above will help us understand the
broad patterning of u €lk€ u-associated violence, its relationship with electoral
uc€
popularization and other variables, this analysis cannot properly explain how
u
€lk€
uc€u movement managed to reconcile its use of violence in a period of discur-
sive moderation or how it managed to avoid both marginalization from the soci-
ety and state repression. To understand these processes, I use qualitative data
from in-depth interviews conducted in seven districts of four cities where Kurd-
ish civilians are frequent victims of communal violence in western Turkey.48
Between 2010 and 2013, I conducted in-depth interviews with 77 local govern-
ment officials, political party leaders/representatives (including the ruling party,
pro-Kurdish party, main opposition party, as well as ultra-nationalist parties
e.g., the MHP), civil society organization representatives (migration institutes,
human rights organizations, labor unions, extreme-right organizations includ-
ing the u u ocakları), and local residents from different ethnic backgrounds. I
€lk€
identified interviewees using snowball sampling. Ranging from 1 to 3 hours, the
interviews were semistructured and open-ended. The interviews included ques-
tions to understand (1) intercommunal economic, political, and social relations
in the neighborhoods and their transformation in time (particularly since the
1990s), (2) different attitudes of residents and representatives of political parties

regarding different forms of collective violence in their neighborhoods, (3) per-
ceptions regarding the causes, mechanisms, actors, and outcomes of societal
nationalist and ethnic violence. The names of locations and interviewees used
throughout this study are pseudonyms.
The fieldwork has allowed me to understand how u €lk€ u militants, other
uc€
political party representatives and local residents perceive the role of u €lk€
uc€
u
movement as well as other political actors in the emergence and spread of collec-
tive violence incidents. Of course, qualitative analysis based on interview data
must be conducted and interpreted very carefully. Recent scholarship points out
the problem of “attitudinal fallacy,” a limitation common to many interview-
based accounts, which arises due to discrepancy between verbal accounts and
actual social action/behavior.49 The analysis presented in this paper uses two
main strategies to avoid “attitudinal fallacy.” First of all, resembling the ethno-
graphic approach, which draws attention to the interpretation of “discrepancies
between saying and doing,”50 the analysis tries to uncover and interpret the dis-
crepancies (1) embedded in the own accounts of the u €lk€ u activists regarding
uc€
movement violence; (2) between accounts of u €lk€
uc€u activists and other actors
involved in the process, as well as (3) between accounts of u €lk€ u activists and
uc€
those found in newspaper data and human rights reports. Unearthing these dis-
crepencies was largely possible since questions I asked to each interviewee was
informed by accounts derived from previous interviews with u €lk€ u activists and
uc€
other actors in the field. This strategy resembles the logic of treating each inter-
viewee as a single case in interview-based studies.51 Second, the analysis does not
take the actual accounts of incidents by u €lk€ u activists with their face value.
uc€
Rather, it uses these accounts to understand the social construction of move-
ment violence, which means how members of the u €lk€ u movement perceive and
uc€
explain utilization of violence.
Findings of the Statistical Analyses
€ uc€
Overview of the Post-1980 Trajectory of the Ulk€ u Movement
Illustrating the temporal patterns of the electoral strength of the MHP and
violent incidents associated with the u €lk€ u movement from 1981 to 2011, Fig-
uc€
ure 2 above has already showed that political violence has not ceased to be a part
of the action repertoire of the u
€lk€
uc€u movement after the 1980s. On the contrary,
the number of violent incidents associated with the u €lk€ u movement increased
uc€
with the transition from the military junta regime to democratization, accompa-
nying the electoral popularization and “moderation” of the MHP. This rising
trend of u€lk€ u-associated political violence came to a peak in 1998.52 Interest-
uc€
ingly, the rising trend of u€lk€
uc€u violence was reversed in 1999, when the MHP
experienced its historic electoral success (both in national and local elections)
and joined the coalition government. Between 1999 and 2002, when the MHP
was a coalition partner of the government, the number of violent incidents asso-
ciated with the u€lk€ u movement remained low. While there was a brief decline
uc€
in levels of violence in the early 2000s, u
€lk€ u violence started to rise once again
uc€
during the AKP era—marked by the “Kurdish opening process” where the gov-
ernment aimed to partially expand democratic rights and liberties of the Kurdish
population to co-opt rising Kurdish movement—and has followed a consistently
rising trend with higher absolute levels compared to any other period in the
post-1980 era of extreme right politics.
The dramatic rise of u€lk€ u violence at the turn of the twenty-first century is
uc€
not independent from the escalation of the Kurdish movement in the same
period. Actually, one of the interesting aspects of the post-1980 trajectory of
u
€lk€ u violence is the change in its targets over time. As a whole, from 1981 to
uc€
2011, 43 percent of all u €lk€ u violence is directed against leftist groups (which
uc€
includes socialists, communists as well as Kurds who are perceived as leftists),
and 27 percent is directed against Kurds and Kurdish activists. While violence
directed against leftist groups is a continuation of the pre-1980 period, violence
against the Kurds is a novelty of this new era. As Figure 5 shows, in every decade,
violence against Kurds gradually increased: It increased from 0 percent in
1981–1991 period to 20 percent in 1991–2001 period, and to 38 percent in the
2001–2011 period. As Figure 6 shows, in the post-2004 period, the era of the
€ uc€
Figure 5. Targets of Ulk€ u-Associated Violence, 1981-Present.
Source: ENViT database.
€ uc€
Figure 6. Targets of Ulk€ u-Associated Violence, 2004-2011.
Source: ENViT database.
“Kurdish opening” led by the AKP government, the frequency of violence

directed against Kurds (41 percent) had already surpassed violence directed at
leftist groups (36.46 percent).
It is not coincidental that in the same period, the MHP also made the
Kurdish problem its primary political propaganda. In the parliament, the MHP
represented itself as the only political party that will not make any concession to
Kurdish political actors (neither the PKK nor legal pro-Kurdish parties) but will
fight against these internal enemies at whatever cost.
The Relation between Far Right Violence and Electoral Strength
There is statistical evidence suggesting a positive association between elec-

toral popularization of the MHP and the use of violence by the u €lk€ u movement
uc€
in post-1980 Turkey. Table 2 shows negative binomial coefficients from the
regression analysis of the number of u €lk€ u-associated violent incidents on
uc€
selected variables. Model 1 examines the effect of the existence of a military junta
regime—as an indicator of state repression—and this variable is included in all
other models. To examine effects of elections and electoral strength on u €lk€ u-
uc€
associated violence, Model 2 includes a series of political variables including
electoral strength of the MHP, whether or not MHP is in government that year,
and whether or not there is an upcoming election. To examine the effects of
socio-economic competition and deprivation on violence Model 3 includes eco-
nomic variables including the GDP per capita growth, unemployment level, and
wage level. Model 4 is a composite model that combines variables used in
Models 1, 2, and 3. Model 5 includes Kurdish mobilization variable into the
composite model. Since Kurdish mobilization and the MHP’s electoral strength
and strongly correlated to each other (r 5 0.75; see Table 1), to avoid multicolli-
nearity, Model 5 excludes the MHP’s electoral strength from the model.53 All
models include the lagged dependent variable as a control variable. Except for
the upcoming elections variable, all variables are lagged one year.
Table 1. Descriptive Statistics and Correlation Matrix

N Mean s.d. 1 2 3 4 5 6 7 8
€ uc€
1- Ulk€ u associated violence 31 7.09 6.37 1.00
2- Electoral strength of the MHP 31 9.56 4.47 0.66 1.00
3- Military junta period 31 0.22 0.42 20.59 20.61 1.00
4- MHP in government 31 0.12 0.34 20.05 0.40 20.21 1.00
5- Election years 31 0.38 0.59 20.03 0.21 20.24 0.25 1.00
6- Kurdish mobilization 31 31.00 3.13 0.64 0.75 20.85 0.17 0.28 1.00
7- GDP per capita growth 31 0.02 0.04 20.10 20.10 0.10 20.30 20.28 0.05 1.00
8- Wage level (19975100) 31 85.37 19.8 0.36 0.33 20.77 0.01 0.24 0.66 0.01 1.00
9-Unemployment level 31 8.27 1.49 20.12 20.01 0.02 0.36 0.17 20.03 20.52 20.16
All models show that existence of a military junta decreases u€lk€ u-associated
uc€
violence, which supports social movement and extreme right literature that
emphasizes how state repression deters movement violence. This finding is not
surprising since military junta period repressed nearly all forms of political action
including violent and nonviolent forms, and banned pre-coup political organiza-
tions. Right-wing violence resurfaced only with the re-institutionalization of
democratic politics.
€ uc€
Table 2. Ulk€ u Associated Violent Incidents in Turkey, 1981–2011 (Negative Binomial)
Model 1 Model 2 Model 3 Model 4 Model 5

€ uc€
Ulk€ u associated violence (t – 1) 0.0313 20.0298 0.0249 20.0250 20.0145
(0.026) (0.020) (0.025) (0.0186) (0.0150)
Military junta regime (1980–1983) (t – 1) 21.806*** 21.143* 22.397*** 22.191*** 21.1231
(0.498) (0.459) (0.629) (0.497) (0.581)
Electoral strength of the MHP (t – 1) 0.129*** 0.100***
(0.034) (0.030)
MHP in government (1999–2002) (t – 1) 21.317*** 21.223*** 20.778**
(0.339) (0.317) (0.260)
Election years (t 1 1) 0.378** 0.366** 0.487***
(0.141) (0.142) (0.129)
GDP per capita growth (t – 1) 4.663 0.243 22.315
(3.194) (2.382) (2.200)
Change in unemployment level (t – 1) 0.627 20.459 21.293
(1.302) (0.916) (0.854)
Wages (t – 1) 20.0136 20.0131 20.018*
(0.011) (0.007) (0.007)
Kurdish mobilization (t – 1) 1.083***
(0.267)
Constant 1.926*** 1.024*** 3.095** 2.528** 20.0292*
(0.280) (0.305) (1.103) (0.782) (1.155)
Ln(alpha) 21.057* 22.727** 21.289** 23.350* 24.563a
(0.413) (0.940) (0.455) (1.554) (5.061)
Log-likelihood 281.851 271.260 279.894 271.856 269.024
McFadden’s adj. R square 0.063 0.146 0.051 0.106 0.137
Number of cases 30 30 30 30 30
Note: 1p < .10, *p < .05, **p < .01, ***p < .001 (two tailed tests). Numbers in parentheses are standard errors.
a
Because the dispersion term alpha was not statistically significant for Model 5, I replicated this model using
Poisson regression analysis. Signs and significance of the variables are identical to the one presented in Model
5 above.
Results also suggest that MHP’s electoral strength and popularization has
not reduced violent events associated with the u €lk€ u movement, controlling for
uc€
other variables. In contrast to moderation thesis, the effect of the MHP’s electoral
strength and popularization (lagged one year) on u €lk€ u-associated violence is
uc€
positive and significant in all models where this variable is included (Models 2
and 4). Table 2 also shows that there is a statistically significant and positive
effect of upcoming elections on u €lk€ u violence as expected by theorists of elec-
uc€
toral violence and political competition (see Models 2, 4, and 5). Furthermore,
the MHP’s being a government partner—the MHP in government variable—
decreases the level of u €lk€ u-associated violence, which suggests that the u
uc€ €lk€
uc€
u
movement is more likely to utilize violence as a strategy of political opposition;
and less likely to utilize violence when it is a governing party.
While indicators of socio-economic competition and deprivation such as
GDP per capita growth rate and change in unemployment do not have explana-
tory power (see Models 3, 4, and 5), negative coefficients of wages in Models 4
and 5 suggest that declining wages have a positive effect on u €lk€ u-associated
uc€
violence, as expected by socioeconomic competition and deprivation theories.
This effect, however, is not robust. However, due to small number of observa-
tions used in the study, these lack of significant coefficient in Model 3 can be due
to Type-II error, hence effect of declining wages on u €lk€ u movement must still
uc€
be considered as a possible complementary explanation.
Moreover, as Model 5 shows, Kurdish mobilization—a composite variable
that includes both Kurdish armed mobilization and Kurdish democratic/elec-
toral mobilization to assess the perception of the “Kurdish threat”—has a signif-
icant effect on u€lk€
uc€u-associated violence. This effect, which is robust to other
specifications, complements findings in Figures 5 and 6 which show Kurds have
gradually become the new targets of the u €lk€ u movement.
uc€
Summary of the Statistical Analyses
There are two overarching findings that emerge from the statistical analyses
presented above. First one pertains to the relation between elections and vio-
lence. The historical patterning of violence suggests that u €lk€
uc€
u violence that
was repressed by the military resurfaced with democratization through re-
institutionalization of competitive elections. Furthermore, as a whole, Figure 2
and Table 2 show that the relationship between MHP’s electoral popularization
and u€lk€ u-associated violence is not a negative one; but violence rises with the
uc€
growing electoral power of extreme right—even controlling for alternative
explanations. Put differently, neither MHP’s electoral popularization deterred
the movement from eliminating violence from its action repertoires, nor move-
ment violence led to loss of popular support and to state repression.
Second, the descriptive analysis of targets of u €lk€ u violence over time
uc€
reveals the political logic of movement violence. Instead of targeting state insti-
tutions, u
€lk€ u violence, like most right-wing movements, engaged in horizontal
uc€
violence toward social and political groups they consider as internal enemies of
the nation/state. Throughout the 1990s, the movement targeted the leftist
movement that was particularly active in the universities. Throughout the
2000s, that is, the decade of the “Kurdish Opening,” its main target shifted
toward the Kurds concomitant to increasing popularization of the Kurdish
movement in the non-Kurdish cities of Turkey. In both cases, the movement
targeted leftist groups which were politically active and visible.
Altogether, these findings generate some puzzles and questions that need to
be explained. First of all, how did u€lk€ u movement managed to maintain its
uc€
popular support, avoid state repression and further marginalization while utiliz-
ing political violence? Furthermore, how does political logic of target selection
shaped strategies that u€lk€ u movement used to reconcile its violence with elec-
uc€
toral politics? Using fieldwork data on violence that targets Kurdish population
in the 2000s, next section presents and explains mechanisms that enable move-
ment reconciliation of violence and electoral politics.
Explaining the Patterns: Findings from Fieldwork
Framing Movement Violence: Discursive Moderation
Since 1990s, in their declarations and public speeches, both the MHP and
the u
€lk€ u movement are very careful not to be associated with the extremism of
uc€
the 1970s,54 which led to movement marginalization and eventual state repres-
sion with the 1980 coup. This change in strategy- together with surprising elec-
toral popularization of the MHP in the 1990s- led many scholars to argue that
the MHP entered a period of moderation. In addition to the statistical evidence
that shows how u €lk€ u associated violence rose concomitant to MHP votes,
uc€
interviews with u €lk€
uc€u activists also suggest that the movement embraces vio-
lence in their daily political actions. At first sight, this discourse on moderation
seems to contradict the violence of the movement. However, discursive modera-
tion, whereby the movement emphasizes moderation but does not altogether
renounce political violence, becomes a strategic tool for the movement to use
violence while presenting itself less violent than it actually is. After all, the way
that violence utilized and framed/presented is as critical as the utilization of
violence.
Discursive moderation revealed itself in the interviews with u €lk€
uc€u activists.
While activists start the conversation by underlining how they refrain from vio-
lence and focus on nonviolent politics, they continue with giving details of how
they actually engage in actions of violence. For instance, an u €lk€ u activist
uc€
explained to me how they follow their leader Devlet Bahçeli’s call for nonviolent
politics, and refrain from fighting with the Kurds as follows:
Q: What is your attitude towards violence against the Kurds in the city?
A: There is a tension in the city. We tell our activists not to be provoked. I
assure you we fight only once when we need to fight five times. Our president
[Devlet Bahçeli] tells us to sell the guns and buy laptops. We saw that it is a
useful strategy.
Q: So, there is not much violence despite these tensions you mentioned.
A: Of course, there are fights. . .especially in the schools [high schools]. In
the city, ocaks are organized in all high schools. We are there, because there
is no peace there, because the PKK sympathizers are there. What we say is
simple: You stop supporting the PKK, and we stop establishing the ocaks.
Let us all abandon politics. . .The only thing that we – the u €lk€
uc€us - want is
€ u Ocag ı, Male)
peace. . . (Author’s interview with a member of Ulk€
This conversation reveals how u €lk€ u activists perceive and explain their
uc€
movement’s tendency towards “non-violence” in this new period. While this
young activist initially underlines how members of the movement refrain from
violence, he does not denounce violence altogether. On the contrary, he empha-
sizes their nonviolent tendency by underscoring that they “fight only once when
they need[ed] to fıght five times.” For him, the fact that the u €lk€ u militants
uc€
“hold themselves back” from fighting in some instances is proof of their desire
for peace and nonviolence. He also underlines that they are only involved in vio-
lence when it is necessary to do so. Not surprisingly, such framing of movement
violence is largely shaped by the political context. According to his explanation,
it is not their inclination to fight but the very existence of internal enemies, such
as “the PKK sympathizers,” which pushes the u €lk€
uc€u activists to engage in these
fights.
Put differently, u
€lk€ u militants I interviewed consistently presented them-
uc€
selves to me as patient and responsible actors (in contrast to irrational, fringe,
extremist ones) who desire peace but are pushed to “fight” and protect the
nation’s unity because of the growing support for internal enemies in a period of
increasing Kurdish organizations and support for the Kurdish movement. How-
ever, the definition of who were the “internal enemies” and “Kurdish separatists”
was very broad for the u €lk€
uc€u movement. Besides supporting the pro-Kurdish
parties/movement, even speaking Kurdish out loud in public space was seen as
evidence of being a separatist, hence a threat to the unity of the nation. One
young u €lk€
uc€ _
u, in a working class neighborhood in Işler, explained to me how
they expelled Kurdish workers from their neighborhood because they were
“speaking Kurdish as if they were bragging”:
It was two years ago - there were ‘bachelor homes’ where Kurdish workers
lived together. They were speaking Kurdish very loudly. Women could not
walk safely in the neighborhood. . . One evening, we raided their residences
with our 300 brothers [i.e., senior members of the movement]. Our brothers made
a decision and they [Kurdish workers] were dispelled from the
neighborhood. . . Yes, they can speak Kurdish, but they can’t do it as if they
are bragging. (Author’s interview with a member of Ulk€€ u Ocagı, Male)
Since they considered the existence of a large group of young Kurdish bach-
elor workers—speaking Kurdish out loud and not intimidated by the u €lk€
uc€
u
warnings—as a threat to the neighborhood, senior members of the movement
decided and carried out this action, which successfully expelled Kurdish workers.
Another young u €lk€
uc€ _
u activist, who was in charge of the local Ocak in Işler, told
me about another very similar incident, while simultaneously complaining about
how only the u€lk€ us were held accountable for instances beyond their control:
uc€
Q: What is your attitude towards violence against the Kurds in the
neighborhood?
A: We are cautious not to get involved in violent incidents. We want to
work with the masses. . .Unfortunately, we are held responsible for the things
beyond our control. During [nationalist] demonstrations, for instance,
people supporting the AKP and the CHP [the two largest political parties in
Turkey] also make ‘greywolf [bozkurt] signs’ [i.e., the hand-sign of the u €lk€uc€
u
movement]; consequently when there is an attack against the BDP [pro-Kurd-
ish party], we are held accountable.
Q: So you are totally against these violence incidents then. . .
A: Sometimes, mmmh, throwing stones at the BDP [pro-Kurdish party]
buildings, etc. . ..these kinds of things are done when necessary. . . After all, we
are fighting for our homeland [vatan]. . . Sometime ago, for instance, there
was a fight with Kurdish construction workers. They were speaking Kurdish
and some of our friends warned them to stop speaking in Kurdish. Kurdish
workers continued and made the ‘victory sign’ in return. Then, we gathered
together and went to their ‘tents’ [he refers to the camp of the construction
workers, who were temporary migrant workers], there was fifteen of us and
sixty of them. There was a fight and I was stabbed. In the end, these
workers were sent away from the neighborhood.
This is a quintessential example of the contemporary u €lk€ u attitude toward
uc€
violence, and key to understanding the notion of discursive moderation. On the
one hand, u €lk€uc€u activists no longer want to be associated with their historic
“violent” image. Hence, they engage in community work in various working
class neighborhoods and they try to build a large network of nationalist youth.
They also consistently emphasize their inclination toward nonviolent politics.
On the other hand, they give many accounts of their actual involvement in vio-
lence when they consider it necessary to defend the motherland, by providing
detailed justifications for why it was necessary to do so and presenting violence as
a tool that was used as a last resort.
Mobilizing Masses for Violence
Another transformation in the way that u €lk€ u movement uses violence is

uc€
their interaction with the masses in this process. While in the 1970s, the u€lk€
uc€
u
movement formed paramilitary groups trained in commando camps and
engaged in militant violence targeted against revolutionary left, today, they give
particular emphasis to their interactions with ordinary masses and mobilize
them for violence. A common theme of the interviews with members of the
u
€lk€ u movement is their emphasis on the participation of other people, such as
uc€
ordinary residents of the town and people that vote for other conservative-
nationalist or mainstream parties, in these violent events.
For instance, when talking about the 2006 lynching attempts in Kirazlı,
where 7–8 thousand people tried to lynch four Kurdish seasonal workers, an
MHP member in Kirazlı “complained” about how u €lk€ us had to bear responsi-
uc€
bility for incidents that were beyond their control:
Whenever such an incident [e.g., lynching attempt] takes place, u €lk€ us are
uc€
held accountable. For instance, during the [2006] incidents here, I saw a kid
supporting the Felicity Party [Refah Partisi, an Islamist Party], who was
making the bozkurt sign [i.e., the hand-sign of the u
€lk€ u movement, repre-
uc€
senting being an u€lk€ u]. When I asked him why he was doing it, he just
uc€
smiled.
The MHP representative did not really seem to be bothered about these
people who had no organizational links with their movement making the bozkurt
hand-sign. Rather than being distressed, he was smiling and seemed content
about this hand gesture. This is not surprising because one of the key mobiliza-
tional tools within the action repertoire of the u
€lk€ u movement is leading people
uc€
to make bozkurt signs while singing the national anthem or shouting nationalist
slogans during nationalist demonstrations. Larger masses making the bozkurt
hand-signs and shouting u €lk€ u slogans actually symbolizes movement’s capacity
uc€
to lead the masses beyond their formal organizational links. In the 1990s, when
the armed conflict with in the Kurdish region reached its peak, nationalist dem-
onstrations became critical tools for u €lk€ u militants who aimed to extend the
uc€
scope of support for the u€lk€ u movement, to transform people’s perceptions of
uc€
existing enemies and threats against the motherland, and even to influence the
electoral choices of these nationalist-conservative masses.
This form of nationalist mobilization of masses in the 1990s has increasingly
taken violent routes, and in the 2000s u €lk€ u activists have increasingly utilized
uc€
their mobilization skills in gathering crowds for demonstrations and converting
these gatherings into violence. Interviews, in addition to the archival research,
reveal the complex roles u €lk€ u militants have played in this process. Members
uc€
of the u€lk€ u movement played key mediating roles in anti-Kurdish mass vio-
uc€
lence—as “violent specialists,”55 “riot specialists,”56 and “rumor-mongers,”
those who direct and organize violence.57 They bring together a protesting
crowd through production and dissemination of rumors, leading nationalist pro-
cessions and in converting nonviolent protests into violent ones.
Rumor Production. One such role is the production and dissemination of

rumors.58 In most cases, people in these neighborhoods gathered together on
rumors such as “Kurds are attacking Turks,” “Kurds harassed our women,”
“Kurds hung a PKK flag,” “Kurds burnt the Turkish flag” and so on. Ordinary
people that participated in nationalist attacks against Kurds in their neighbor-
hoods often stated that they heard about these instances—that is, rumors about
Kurds’ provocations—from the young u €lk€ us. Sometimes they just mentioned
uc€
them as gençler (“the youth”), but people from different civil society organiza-
tions confirmed that they belonged to the u€lk€ u youth.
uc€
The nature of these rumors differs in time and space. Sometimes, they
include a distorted or selected part of the truth. In Şenk€oy, for instance, a mob
violence against the Kurdish residents in town was triggered by a rumor that
“Kurds attacked Turks,” which was actually a personal discussion between a
Kurdish member of the MHP and a Turkish member of the pro-Kurdish party
BDP (Interview with the Representative of Human Rights Association). Some-
times, however, these rumors seem to be completely fabricated. It is surprising
to see how many mobs and lynching attempts seem to originate out of ordinary,
daily discussions such as speaking Kurdish in public, listening to Kurdish music,
wearing Ahmet Kaya (a political Kurdish musician) t-shirts, and so on. These
sorts of daily discussions by themselves rarely create collective raids and lynch-
ing attempts against the Kurds. These discussions seem to radicalize and bring
together a larger crowd on rumors linking these Kurds’ identities and their activ-
ities with the PKK.
Production of Violence. Another key role played by u €lk€ u militants in these

uc€
larger events is leading the “crowd” for basic tasks in the course of the protest
such as where to go and what to do and converting nonviolent gatherings into
violent ones.59 People who participated in these incidents often mentioned that
they originally went out to the street to “protest” the PKK but followed the rest
of the coordinated youth [the u €lk€
uc€u youth], who led them to shout slogans, sing
national anthems, make bozkurt signs, and later throw stones at buildings and to
raid shops. Put differently, u €lk€
uc€u militants manage to convert a nonviolent
political protest into a violent one in the course of the demonstrations. This role
also has similarities with the role played by “conversion specialists” in Brass’s
study of riots in India, who convert local incidents or public issues into violent
ones by stone throwing, stabbing, and arson.60 A mukhtar61 reveals the leading
role of the u
€lk€ u activists in converting a nationalist demonstration into anti-
uc€
Kurdish riots in Durusu in 2011, where thousands of people raided shops of
Kurdish residents:
Of course at the very front of the events was the MHP. But people were
also there. Women etc., they all joined the events. The youth of the MHP
[i.e., u
€lk€
uc€
u activists] started throwing stones at the buildings [of Kurdish
residents] and shouted, “Here is Turkey!” [not Kurdistan]. People did not
react to the Turkish nationalists. They reacted to the Kurds.
Participation of a crowd—consisting of women and the elderly—beyond the
organizational links of the MHP was also confirmed by local residents, mukhtars,
and representatives of other main political parties (AKP and CHP) whom I
interviewed. In addition to ethnic Turks, residents of Laz origin, Kazakh, and
Afghan immigrants also participated in the anti-Kurdish protests. Some local
residents interpreted this wider participation by the crowd as evidence that these
events were beyond the control of Ulk€ € u Ocakları and the MHP. Still, the
accounts of (Turkish) residents who personally participated in the incidents—

and local shop owners in the neighborhood who witnessed the incidents—it was
the young u€lk€ u nationalists who first started throwing stones at Kurdish shops
uc€
and soon everyone followed. A Kurdish coffee shop owner, whose shop was
attacked by the nationalist crowd, points out the leadership of the u€lk€ u youth
uc€
within the crowd.
_
My windows were broken while people started clapping and chanting Istiklal
Marşı [the national anthem]. They were led by an u
€lk€ u group, a group that
uc€
calls themselves as idealists. It was a young group composed of teenagers.
They even shouted as takbir when breaking windows. And they are supposed
to be Muslims. . . Even Kazakh and Afghan people followed them. (Kurdish
Coffee-Shop Owner, Durusu).
Overall, for u
€lk€ u militants, participation of larger masses in these violent
uc€
incidents—especially participation by people supporting other parties—is evi-
dence that the u €lk€ u movement—alone—cannot not be held accountable for
uc€
these violent events and that these events are legitimate—as opposed to being
extremist or fringe—because they reflect popular reactions of the Turkish nation
as a whole. Put differently, mobilizing crowds for violence not only creates the
popular legitimacy for violence but also enables the movement to displace
responsibility and blame on and blame on other actors, including the “crowd”
and the “masses.” This echoes Wilkinson’s note on communal violence in India:
“[i]f one demonstrator throws a stone, it is interpreted as ‘the crowd’ throwing
stones.”62
Interactions with the State
The u€lk€ u violence in the post-coup period also differs from typical forms
uc€
of fascist and far right violence—such as anti-leftist u €lk€ u violence of the
uc€
1970s—whose indirect aim is to destabilize the state63 and which encroaches
upon the state’s monopoly over the use of violence. Interviews suggest that in
the recent period, u€lk€ u movement is very careful not to disrupt public order
uc€
and emerge as a threat to the political establishment in a way that would create a
backlash and trigger repression of the movement by the state, akin to the one in
the post-1980 coup period.
To enable this, members of the u €lk€ u movement play a dual role by simulta-
uc€
neously triggering violent events as well as trying to stop them. Put differently,
members of the u €lk€ u movement not only gather the lynch mobs, they also
uc€
“disperse” them in collaboration with the police. To give an example of this
dynamic, we can have a closer look at one such case that occurred in Kirazlı in
2006. A discussion between four young Kurdish seasonal workers and a local
resident of Turkish origin (who was known to be an u €lk€ u, according to the in-
uc€
depth interviews conducted in the town and reports by human rights organiza-
tions describing the incident) in a grocery store suddenly sparked a large-scale
lynching attempt in which thousands of people participated.
The discussion was about workers’ speaking Kurdish out loud in the grocery
store. After the discussion, u€lk€ u activists spread a rumor in a nearby park in
uc€
order to gather a lynch mob, saying that Kurds opened the PKK flag and fired
guns. On the spread of the rumor in the town, thousands of people gathered
together to punish the “PKK militants.” To escape the gathering crowd, Kurd-
ish workers took refuge in the police station. The crowd came in front of the sta-
tion, and got bigger as people from neighboring villages and towns also joined.
Around 7–8 thousand people stayed in front of the police station until 3 am,
shouting that they will take the Kurdish boys away from the police and punish
them. Interviewees who witnessed the events in the town emphasized the exis-
tence and leadership of the u €lk€ u activists within the crowd, which was in
uc€
accordance with reports of two different human rights institutions. The MHP
representatives I interviewed also admitted the existence of young u €lk€ u acti-
uc€
vists in the events but they emphasized how they actually tried to disperse these
young u €lk€
uc€us and to placate the masses. As one MHP representative explained
to me:
There were 10–15 people in front of the police station in the beginning [of
the events]. When I went there, I saw 2–3 kids from the ocaks [he refers to
young members of the Ulk€€ u Ocakları]. When the police told me to disperse
the kids, I warned them [the u €lk€
uc€ u youth]. But soon, a larger crowd started
to gather in front of the station. . . And the people who were there were not
merely u€lk€ us, they were conservative-nationalist masses. In order to stop
uc€
the event, we started to sing the national anthem and sang it five times. . .
No one besides us cooperated with the police. We [together with the police]
stopped people from entering the police station.
The seemingly contradictory explanations about the role that right wing
political actors played in these kinds of mobs and lynching attempts—both start-
ing the mob violence and trying to stop it—suggest the existence of an implicit
form of division of labor between young u €lk€ u activists and senior MHP mem-
uc€
bers. While young u €lk€ u activists mobilize masses for the “protection” of their
uc€
neighborhoods, villages and cities from “separatists” on the street, the senior
MHP members emerge as responsible leaders of the crowd who make sure that
these events do not go out of control. In the conversation mentioned above, it
was noteworthy to observe the MHP representative’s sensitivity towards acting
in accordance with the police, trying to disperse the young u €lk€ us when police
uc€
asked him to and his emphasis on cooperating with the police to stop the
“people” from lynching the Kurds.
While young u €lk€ u activists emerge as representatives and mobilizers of
uc€
“nationalist sentiments” of the Turkish nation in the eyes of the nationalist con-
servative masses, who warn them of and mobilize them against agitations of
“internal enemies,” the MHP members present themselves as rational, patient,
and responsible actors in the eyes of the state institutions who understand and
share the “nationalist” sensitivities of these people but choose to cooperate with
the state so as not to create chaos. Through this double strategy, the broader
movement expands its mobilizing (both violent and nonviolent), organizing and
leadership capacity over the nationalist-conservative masses in these neighbor-
hoods without triggering state repression.
Conclusion
While dominant perspectives in the literature on far right politics and social
movements point out the incompatibility between the use of violence and elec-
toral popularization, in recent decades, right-wing political movements that use
violence have increased their political salience in various democracies. Existing
research has shed light into relation between votes and violence in cases such as
India and Greece have provided insights on how electoral mobilization, political
competition, and systemic crises have enabled political parties with violent
action repertoires to become central figures in their respective political systems.
The analysis of another negative case, the Turkish extreme-right, brings
some additional light on the interrelated processes of political contexts and
movement strategies, which help extreme right movements to avoid state repres-
sion and marginalization despite their use of political violence. Through an anal-
ysis of qualitative data, the paper presented how u €lk€ u movement crafted the
uc€
following strategies to avoid marginalization and state repression in a political
context of heightened Kurdish mobilization: (1) framing of movement violence
through discursive moderation (2) mobilizing crowds for violence through inter-
action with masses (3) not directly challenging state institutions. Overall, rather
than movement rigidity, which is generally used to describe violent/extremist
political actors, it was this strategic flexibility and innovative capacity that enabled
the u€lk€
uc€u movement to successfully utilize violence with electoral politics in the
last two decades. This strategic flexibility eventually enabled the u €lk€ u move-
uc€
ment to fundamentally alter the repertoire of political violence in Turkey in the
recent decades, manifested by the rise and normalization of civilian/communal
forms of political violence. In the current context of rising political and geopolit-
ical crises in Turkey, far right groups find increasing opportunities to escalate
both the intensity and the severity of this violence. While it is beyond the scope
of this paper, further research reveals that some of these strategies have also been
emulated by Turkey’s governing party—the AKP—in recent years. Establish-
ment of the Osmanli Ocaklari (which broadly follows the model of the Ulk€ € u
Ocakları), the AKP’s increasing reliance on violence to maintain and advance its
popular vote, and efforts by the AKP to mobilize the MHP electorate in the
political sphere are some examples of this emulation.64
Rather than trying to generalize from a single case of extreme right move-
ment, this paper focused on Turkish extreme right as a negative case study which
extends our understanding of simultaneous rise of extreme right votes and vio-
lence by highlighting some of the gaps in the literature. In order to explain why
we have been observing a simultaneous rise of violent populism in different parts
of the globe in recent years, future research should focus on systematic compari-
sons of cases where far right votes and violence simultaneously rise in the Global
South (i.e., India, Greece, and Turkey) as well as in the Global North. The
results of the analysis presented from this study suggests that such a comparative
analysis should also take into account specific political contexts and conjunctures
(in addition to economic contexts and crises) which might help right wing
groups and parties to maintain and even increase their electoral appeal while
their sustained use of political violence, as well as innovative strategies used by
new far right movements around the world which might help them avoid repres-
sion by the state and marginalization by the masses.
Sefika Kumral is a Post-Doctoral Researcher at the Department of Sociology at

Johns Hopkins University. Her main areas of research include comparative-
historical and political sociology with a focus on ethnic violence, democratiza-
tion, social movements, and far-right politics. Her dissertation, “Democracy
and Ethnic Violence: Societal Origins of Anti-Kurdish Communal Violence in
Turkey” (2017, Johns Hopkins University) examines the emergence of anti-
Kurdish communal violence in Turkey in the twenty-first century. She pub-
lished articles and book chapters on historical fascism, militarism, development,
and global social unrest.
Notes
I would like to thank to Beverly J. Silver, Joel Andreas, Ho-Fung Hung, Andreas Wimmer, Rory M. McVeigh,
Sahan Savas Karatasli, and Smriti Upadhyay for their comments on earlier drafts of this paper.
1. Martin Schain, Aristide Zolberg, and Patrick Hossay, Shadows over Europe: The Development and Impact of
the Extreme Right in Western Europe (New York: Palgrave Macmillan, 2002); Kai Arzheimer, “Working-
Class Parties 2.0? Competition between Centre-Left and Extreme Right Parties,” in Class Politics and the
Radical Right, ed. J. Rydgren (New York: Routledge, 2013). 75–90, 75.
2. Hans-Georg Betz, Radical Right Wing Populism in Western Europe (New York: St. Martin’s Press, 1994);
Pippa Norris, Radical Right: Voters and Parties in the Electoral Market, (Cambrdige: Cambridge University
Press, 2005); Cas Mudde, The Populist Radical Right in Europe (Cambridge: Cambridge University Press,
2000).
3. Antonis A. Ellinas, “The Rise of the Golden Dawm: The New Face of the Far Right in Greece,” South
European Society and Politics, (2013) 18(4): 543–65; Adam LeBor, “Marching Back to the Future: Magyar
Garda and the Resurgence of the Right in Hungary,” Dissent 55(2) (2008): 34–8.
4. Michael Johns, The New Minorities of Europe: Social Cohesion in the European Union (Lanham, Maryland:
Lexington Books, 2014).
5. Tore Bjo€rgo and Rob Witte, Racist Violence in Europe (New York, NY: St. Martin’s Press, 1993); Thomas
F. Pettigrew, “Reactions Toward the New Minorities of Western Europe,” Annual Review of Sociology
(1998) 24: 77–103; Cas Mudde, “Central and Eastern Europe,” in Racist Extremism in Central and Eastern
Europe, ed. Cas Mudde (Oxon: Routledge, 2005), 247–64.
6. Hainsworth Paul, The Extreme Right in Western Europe (London: Routledge, 2008); Marco Giugni,
Ruud Koopmans, Florence Passy, and Paul Statham, “Institutional and Discursive Opportunities for
Extreme-Right Mobilization in Five Countries,” Mobilization: An International Journal (2005) 10(1):
145–62, 146.
7. Ruud Koopmans, “Explaining the Rise of Racist and Extreme Right Violence in Western Europe: Griev-
ances or Opportunities,” European Journal of Political Research (1996) 30: 185–218.
8. Mabel Berezin, Illiberal Politics in Neoliberal Times: Culture, Security, and Populism in the New Europe (Cam-
bridge: Cambridge University Press, 2009), 10.
9. Ignacio Sanchez-Cuenca and Paloma Aguilar, “Terrorist Violence and Popular Mobilization: The Case of
the Spanish Transition to Democracy,” Politics & Society (2009) 37(3): 428–53, 429; Donatella della Porta
and Sidney Tarrow, “Unwanted Children: Political Violence and the Cycle of Protest in Italy: 1966–
1973,” European Journal of Political Research (1986) 14: 607–32; Sidney G. Tarrow, Democracy and Disorder :
Protest and Politics in Italy, 1965–1975 (Oxford: Oxford University Press, 1989); Mark R. Beissinger, Nation-
alist Mobilization and the Collapse of the Soviet State (Cambridge: Cambridge University Press, 2002).
10. Donatella della Porta and Sidney Tarrow, Unwanted Children: Political Violence and the Cycle of Protest in
Italy: 1966–1973, European Journal of Political Research (1986) 14: 607–632; Ignacio Sanchez-Cuenca
and Paloma Aguilar, Terrorist Violence and Popular Mobilization: The Case of the Spanish Transition to Democ-
racy, Politics & Society (2009) 37(3): 428–453.
11. Will H. Moore, “Repression and Dissent: Substitution, Context, and Timing,” American Journal of Political
Science (1998) 42: 851–74; Mark Irving Lichbach, “Deterrence or Escalation? The Puzzle of Aggregate
Studies of Repression and Dissent,” The Journal of Conflict Resolution (1987) 31: 266–97.
12. Ruud Koopmans, “The Dynamics of Protest Waves: West Germany, 1965 to 1989,” American Sociological
Review (1993) 58: 637–58.
13. Paul R. Brass, The Production of Hindu-Muslim Violence in Contemporary India (Seattle, Washington:
University of Washington Press, 2003); Steven I. Wilkinson, Votes and Violence: Electoral Competition and
Ethnic Riots in India (Cambridge: Cambridge University Press, 2004); Raheel Dhattiwala and Michael
Biggs, “The Political Logic of Ethnic Violence: The Anti-Muslim Pogrom in Gujarat, 2002,” Politics &
Society (2012) 40(4): 483–516.
14. Antonis A. Ellinas, “The Rise of the Golden Dawn: The New Face of the Far Right in Greece,” South
European Society and Politics (2013) 18(4): 543–65; Geoff, Eley, “What Produces Fascism: Preindustrial
Traditions or a Crisis of a Capitalist State,” Politics & Society (1983): 12(1): 53–82; Philip Morgan, Italian
Fascism, 1915–1945 (Palgrave Macmillan, 2003).
15. Karl Polanyi, The Great Transformation: The Political and Economic Origins of Our Times (Boston: Beacon
Press, [1944] 2001); Gert Krell, Hans Nicklas, and Anne Ostermann, “Immigration, Asylum, and Anti-
Foreigner Violence in Germany,” Journal of Peace Research (1996) 33: 153–70; Lauren M. McLaren,
“Explaining Right-Wing Violence in Germany: A Time Series Analysis,” Social Science Quarterly (1999):
80(1): 166–80.
16. Fredrik Barth, Ethnic Groups and Boundaries: The Social Organization of Culture Difference (London: Allen &
Unwin, 1969); Susan Olzak, The Dynamics of Ethnic Competition and Conflict (Stanford: Stanford University
Press, 1992).
17. Wilkinson, Votes and Violence: Electoral Competition and Ethnic Riots in India; Dhattiwala and Biggs, The
Political Logic of Ethnic Violence: The Anti-Muslim Pogrom in Gujarat, 2002.
18. Wilkinson, Votes and Violence: Electoral Competition and Ethnic Riots in India.
19. Juan Linz, “Some Notes Toward a Comparative Study of Fascism in Sociological Historical Perspective,”
in Fascism: A Readers’ Guide: Analysis, Interpretations, Bibliography, ed. Walter Laqueur (Berkeley: University
of California Press, 1978), 12.
20. Michael Mann, Fascists (Cambridge: Cambridge University Press, 2004), 16.
21. Alexandra Koronaiau, Evangelos Lagos, Alexandros Sakellariou, Stelios Kymionis, and Irini
Chiotaki-Poulou, “Golden Dawn, Austerity and Young People: The Rise of Fascist Extremism
Among Young People in Contemporary Greek Society,” The Sociological Review (2015) 63, S2:
231–49, 231.
22. Rebecca Jean Emigh, “The Power of Negative Thinking: The Use of Negative Case Methodology in the
Development of Sociological Theory,” Theory and Society (1997) 26(5): 649–84; also see, Dylan Riley,
“Civic Associations and Authoritarian Regimes in Interwar Europe: Italy and Spain in Comparative
Perspective,” American Sociological Review (2005) 70(2): 288–310.
23. Feroz Ahmad, The Making of Modern Turkey (London and New York: Routledge, 1993), 144a.
24. Jacob M. Landau, “The Nationalist Action Party in Turkey,” Journal of Contemporary History (1982) 17(4):
587–606, 597.
25. Hugh Poulton, Top Hat, Grey Wolf and Crescent: Turkish Nationalism and the Turkish Republic (London:
Hurst & Company, 1997), 60.
26. Jacob M. Landau, The Nationalist Action Party in Turkey, 594–595.

€
27. Ziya Oniş, “Globalization, Democratization and the Far Right: Turkey’s Nationalist Action Party in
Critical Perspective,” Democratization (2003) 10(1): 27–52.
28. Hasan Ersel, Ahmet Kuyas, Ahmet Oktay, and Mete Tuncay, Cumhuriyet Ansiklopedisi 1923–2000: Cilt IV,
1981–2000 (Istanbul: YKY Yayinlari, 2002), 10.
29. M. Hakan Yavuz, “The Politics of Fear: The Rise of the Nationalist Action Party (MHP) in Turkey,”
Middle East Journal (2002) 56(2): 200–21; Bulent Aras and Gokhan Bacik, “The Rise of Nationalist Action
Party and Turkish Politics,” Nationalism and Ethnic Politics (2007) 6(4): 48–64; Alev Çınar and Burak
Arıkan, “The Nationalist Action Party: Representing the State, the Nation or the Nationalists?” Turkish
Studies (2002) 3(1): 25–40; Nergis Canefe and Tanıl Bora, “The Intellectual Roots of Anti-European
Sentiments in Turkish Politics: The Case of Radical Turkish Nationalism,” Turkish Studies (2003) 4(1):
127–48.
30. Although ban on political activity was removed in 1983 with the return to the civilian rule, the ban on pre-
coup political parties including the MHP remained intact while hundreds of politicians were also barred
from politics until 1987. Due to the political ban on pre-coup political parties, the MHP was re-established
under different names: Muhafazakar Parti (Conservative Party, MP) in 1983, Milliyetçi Çalışma Partisi
(Nationalist Task Party, MÇP) in 1985, and finally again – as the Milliyetçi Hareket Partisi (Nationalist
Action Party) in 1993.
31. See Wilkinson, Votes and Violence: Electoral Competition and Ethnic Riots in India.
32. For similar treatments of data see Matthew Desmond, “Eviction and the Reproduction of Urban Poverty,”
American Journal of Sociology (2012) 118(1) 88–133; Terence K. Hopkins, “World-Systems Analysis: Meth-
odological Issues,” in World-Systems Analysis: Theory and Methodology, ed. Terence K. Hopkins and Imma-
nuel Wallerstein (Beverly Hills: Sage Publications, 1982), 145–58.
33. In the construction of the ENViT database, Milliyet is selected as the main source because of its balanced
coverage of the subject matter of interest and higher quality historical archives. Preliminary research
showed that compared to Milliyet, other major national newspapers like Radikal or Cumhuriyet had wider
coverage of right-wing nationalist violence. Yet in these center-left wing newspapers there was a higher
tendency to use terms “fascists” or “€ ulk€uc€
u” as generic terms for perpetrators of any kind of incidence of
violence against workers and ethnic/religious minorities. Hence, these center-left oriented newspapers are
not preferred in order to minimize selection, description, and data collection biases. Furthermore, com-
pared to H€ urriyet, which is the other major mainstream newspaper, Milliyet was preferable because of its
higher quality digital historical archive, which helps minimize “data collection bias.” While selection of
Milliyet as the main source resulted in a more conservative estimate of the total number of incidents, its
coverage of extreme right violence incidents is high enough to capture most major events, and thanks to
the quality of the historical archives, this coverage does not have a systematic bias across time. Reliability
studies based on a comparison of the ENViT database—using Milliyet as its main source—to the existing
secondary literature and Human Rights Association reports show that the ENViT database does not miss
any of the major incidents and it does not have a systematic temporal bias.
34. Jeffery M. Paige, Agrarian Revolution (New York: Free Press, 1975); David Snyder and Charles Tilly,
“Hardship and Collective Violence in France, 1830 to 1960,” American Sociological Review (1972) 37(5):
520–32; Kriesi et al., New Social Movements in Western Europe: A Comparative Perspective; Roberto Franzosi,
“Computer-Assisted Content Analysis of Newspapers.” Quality and Quantity (1995) 29(2): 157–72.
35. Olzak, The Political Context of Competition: Lynching and Urban Racial Violence, 1882–1914; Ashutosh Var-
shney, Mohammad Zulfan Tadjoeddin, and Rizal Panggabean, “Creating Datasets in Information-Poor
Environments: Patterns of Collective Violence in Indonesia, 1990–2003,” Journal of East Asian Studies
(2008) 8(3): 361–94; Wilkinson, Votes and Violence: Electoral Competition and Ethnic Riots in India.
36. Ruud Koopmans and Susan Olzak, “Discursive Opportunities and the Evolution of Right-Wing Violence
in Germany,” American Journal of Sociology (2004) 110(1): 198–230; Donatella della Porta, Manuela Caiani,
and Claudius Wagemann. Mobilizing on the Extreme Right: Germany, Italy, and the United States (Oxford
Scholarship Online, 2012).
37. Beverly J. Silver, Forces of Labor: Workers’ Movements and Globalization since 1870 (Cambridge: Cambridge
University Press, 2003).
38. “Selection bias” emerges due to newspapers not reporting all events or reporting them selectively with a
systematic bias over space and time, and (2) the “description bias” emerges due to missing or incorrect
information in news reports, and (3) the “data collection bias” emerges due to systematic biases that
emerge because of data collection schemes employed. See, Jennifer Earl, Andrew Martin, John D. McCar-
thy, and Sarah A. Soule, “The Use of Newspaper Data in the Study of Collective Action.” Annual Review of
Sociology (2004): 65–80.
39. Patrick Barron Sana Jaffrey, Blair Palmer, and Ashutosh Varshney, “Final Reflections: Looking Back,
Moving Forward, in Collective Violence in Indonesia, ed. Ashutosh Varshney (Boulder: Lynne Rienner 2008);
Ashutosh Varshney, Mohammad Zulfan Tadjoeddin, and Rizal Panggabean, Creating Datasets in
Information-Poor Environments: Patterns of Collective Violence in Indonesia, 1990–2003.
40. Susan Olzak, “The Political Context of Competition: Lynching and Urban Racial Violence, 1882–1914,”
Social Forces (1990) 69(2): 395–421; Wilkinson, Votes and Violence: Electoral Competition and Ethnic Riots in
India; Brass, The Production of Hindu-Muslim Violence in Contemporary India.
41. All variables are lagged one year. Unemployment data is from the IMF estimates. All other economic varia-
bles and the data on general and local elections come from TUIK (http://tuikapp.tuik.gov.tr). Data on
military deaths due to armed conflict is from Nedim Şener, “26 Yılın Kanlı Bilançosu,” Milliyet (June 24,
2010), http://www.milliyet.com.tr/26-yilin-kanli-bilancosu-gundem-1254711/; Tolga Şardan, “28 Yılın
Acı Bilançosu: 35 Bin 300 Kişi Ter€or Kurbanı Oldu,” Milliyet (August 16, 2012), http://siyaset.milliyet.
com.tr/28-yilin-aci-bilancosu-35-bin-300-kisi-teror-kurbanioldu/siyaset/siyasetdetay/16.08.2012/1581690/
default.htm. In the preliminary analysis, I also considered rate of unemployment, number of forced Kurdish
immigrants to Western cities and percentage of Kurdish population in internal migration in Turkey as possi-
ble alternative measures. The findings were not robust and they did not alter the existing findings, hence I
did not use these variables.
42. Hanspeter Kriesi, Ruud Koopmans, Jan Willem Duyvendak, and Marco Giugni, New Social Movements in
Western Europe: A Comparative Perspective (Minneapolis: University of Minnesota Press, 1995); also see
Koopmans, Explaining the Rise of Racist and Extreme Right Violence in Western Europe: Grievances or
Opportunities.
43. In preliminary analysis, I also considered the natural logarithm of population as a control variable. Since
the dependent variable is a count of u €lk€ u-associated violent events, it is likely to depend on the popula-
uc€
tion size that year. While the addition of the control variable did not alter the results present/ed in this arti-
cle, addition of this control variable created multicollinearity problem because of its high correlation with
the Kurdish electoral success variable.
44. See Wilkinson, Votes and Violence: Electoral Competition and Ethnic Riots in India; Beissinger, Nationalist
Mobilization and the Collapse of the Soviet State.
45. We lose another observation because of the lagged data. Hence total number of observations is 30.
46. Olzak, The Political Context of Competition: Lynching and Urban Racial Violence, 1882–1914; Beissinger,
Nationalist Mobilization and the Collapse of the Soviet State.
47. In addition, considering higher possibility of Type II errors, we also included p < .10 into the regression
models.
48. As I will describe in more detail below, one of the most interesting aspects of u €lk€ u-associated violence in
uc€
the last decade is that its targets are increasingly becoming Kurdish residents living in Western cities. Since
2004, 41.76 percent of all targets of u €lk€ u-associated violence were Kurdish population.
uc€
49. Colin Jerolmack and Shamus Khan, “Talk is Cheap: Ethnography and the Attitudinal Fallacy,” Sociological
Methods & Research (2014): 1–32, 2.
50. Ibid, 5.
51. See Mario Luis Small, “‘How many cases do I need?’ On Science and the Logic of Case Selection in
Field-Based Research,” Ethnography (2009) 10(1): 5–38.
52. This was an unusual period whereby three simultaneous developments coincided: First, there was rising
left-wing militancy in the 1995–1999 period which started with the Gazi uprising of 1995 and the historic
May-Day of 1996, marking the revival of the socialist movements in the post-coup period. This revival was
mostly felt in university campuses and intensified the clashes between the u €lk€ u students and leftists
uc€
(including left wing Kurdish youth) in university campuses. Second, in July 1998, an explosion in the his-
toric Spice Bazaar (Mısır Çarşısı) in Istanbul, started a serious of lynching attempts to “suspects” of the trial
(including the sociologist Pinar Selek) all of whom were thought to be linked with the PKK. Third, in
October–November 1998, Turkish government pressured Syria to force the PKK leader Abdullah Ocalan €
€
to leave the country and started a process of chasing and capturing Ocalan.
53. Correlation between Kurdish mobilization and the military junta is also very strong (r 5 20.85) but
because military junta is a dichotomous variable, it results with a smaller variance inflation factor score
(VIF 5 6.13). So in this model, I did not exclude the military junta variable. Yet, further analysis shows that
exclusion of this variable from the model does not affect the results presented in Model 5. Likewise, with-
out controlling for other variables, Kurdish mobilization variable is still significant at 0.001 level.
54. Devlet Bahçeli’s response to u
€lk€ u militants shouting “we will kill if you command; we will die if you
uc€
command!” is “Wait! Its time will come!” T24, Bahçeli: ’Vurmanın da zamanı gelecek’ s€ ozlerimin
arkasındayım! 03 26. http://t24.com.tr/haber/bahceli-vurmanin-da-zamani-gelecek-sozlerimin-arkasin-
dayim,226483.
55. Charles Tilly, The Politics of Collective Violence (Cambridge: Cambridge University Press, 2003).
56. Brass, The Production of Hindu-Muslim Violence in Contemporary India.
57. Stanley J. Tambiah, Leveling Crowds: Ethnonationalist Conflicts and Collective Violence in South Asia (Berkeley
and Los Angeles, CA: University of California Press, 1996).
58. Brass, The Production of Hindu-Muslim Violence in Contemporary India, 32–33; Also see Tanıl Bora,
T€ _
urkiye’nin Linç Rejimi (Istanbul: _
Iletişim Yayınları, 2008); Zeynep Gambetti, “Linç Girişimleri,
Neo-Liberalizm ve G€ uvenlik Devleti,” Toplum ve Bilim (2007) 109: 7–34.
59. As Tilly notes, “most collective violence –in the sense of interactions which produce direct damage to per-
sons and objects—grows out of actions which are not intrinsically violent,” See Charles Tilly, From
Mobilization to Revolution (Reading, Massachusetts; Menlo Park, California; Don Mills, Ontario:
Addison-Wesley Publishing Company, 1978), 177.
60. Brass, The Production of Hindu-Muslim Violence in Contemporary India, 33.
61. Elected neighborhood/village headman. Mukhtars has a unique position in the governing structure of
Turkey, providing the link between citizens in neighborhoods with the state.
62. Wilkinson, Votes and Violence: Electoral Competition and Ethnic Riots in India, 24.
63. See Mann, Fascists.
64. Sefika Kumral, “Democracy, Crisis and Geopolitics: Anti-Kurdish Riots and the AKP’s Authoritarian
Turn in Turkey.” In Albert Bergesen and Christian Suter (eds.) Return of Geopolitics (New York: LitVer-
lag, forthcoming).

Ballots With Bullets - The Rise of The Extreme Right in Turkey

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ballots With Bullets - The Rise of The Extreme Right in Turkey

Uploaded by

Copyright:

Available Formats

LANDS AND

BALLOTS WITH BULLETS: ELECTIONS, VIOLENCE,

The rise of extreme right parties and movements in liberal democracies in

violent Islamophobic actions utilized by extreme right organizations such as

The Negative Case of Turkish Far Right

Figure 3. Historical Trajectory of Far Right in Turkey: Two Perspectives.

Data and Methods

This paper uses a combination of quantitative and qualitative analysis to

In the existing literature on far right movements in Turkey, there is no exist-

collective violence incidents associated with the u€lk€ u movement, is filtered

Independent Variables. According to the existing “elections-violence dualism”

mobilization” captures these two distinct dimensions as well as the transforma-

space (e.g., cities and regions). Conducting a cross-sectional time-series or

1990s), (2) different attitudes of residents and representatives of political parties

Findings of the Statistical Analyses

“Kurdish opening” led by the AKP government, the frequency of violence

The Relation between Far Right Violence and Electoral Strength

There is statistical evidence suggesting a positive association between elec-

Table 1. Descriptive Statistics and Correlation Matrix

Model 1 Model 2 Model 3 Model 4 Model 5

Summary of the Statistical Analyses

Explaining the Patterns: Findings from Fieldwork

Framing Movement Violence: Discursive Moderation

Mobilizing Masses for Violence

Another transformation in the way that u €lk€ u movement uses violence is

Rumor Production. One such role is the production and dissemination of

Production of Violence. Another key role played by u €lk€ u militants in these

accounts of (Turkish) residents who personally participated in the incidents—

Interactions with the State

Sefika Kumral is a Post-Doctoral Researcher at the Department of Sociology at

26. Jacob M. Landau, The Nationalist Action Party in Turkey, 594–595.

You might also like