bs_bs_banner

Political Psychology, Vol. 35, No. 1, 2014

doi: 10.1111/pops.12167

Presidential Address
Identity Politics and Politicized Identities: Identity Processes and
the Dynamics of Protest
P. G. Klandermans
VU-University

Over the last decades, the concept of identity has become increasingly central in the social psychology of
protest. Collective identity, politicized collective identity, dual identity, and multiple identities are concepts that
help to understand and describe the social psychological dynamics of protest. In this article, I theorize about
identity processes in the context of protest participation: how group identification establishes the link between
social identity and collective identity, how multiple identities and dual identities influence protest participation,
and how collective identity politicizes and radicalizes. I will illustrate my argument with results from research
into collective action participation among farmers in the Netherlands and Spain, Turkish, and Moroccan
immigrants in the Netherlands and New York, South African citizens, and participants in street demonstrations
conducted by my research group at VU-University.
KEY WORDS: collective identity, group identification, social movements, dual identity, protest participation, radicalization

Two years ago, the Dutch government proposed to cut back on financial support of the disabled.
Within weeks, a mass demonstration of disabled citizens and their caregivers took place in The
Hague—the seat of the country’s government—to protest against the austerity measures. Identity
politics had bred politicized identities. Politics is about the distribution of goods and bads in society.
The allocation process almost inevitably turns into identity politics even if governments painstak-
ingly try to not advantage or disadvantage the one group over the other. But, frequently governments
fail to secure equality in society or deliberately establish and maintain inequality. Think of the Blacks
under the apartheid’s regime in South Africa, Arabic citizens in Israel, or immigrants in the
Netherlands. Identity politics makes collective identities salient, contributes to their politicization
and radicalization, and triggers political protest.
Identity processes play a crucial role in the dynamics of protest (van Stekelenburg, 2013)—as
antecedents, mediators, moderators, or consequences. Yet, identity did not always feature promi-
nently in the social or political psychology of protest, if only, because social psychologists did not
pay much attention to political protest to begin with. This has now changed—a small but growing
contingent of social psychologists is involved in studies of protest behavior, and in their models, the
concept of identity occupies a central place (Azzi, Chryssochoou, Klandermans, & Simon, 2011; Iyer
& van Zomeren, 2009; Stryker, Owens, & White, 2000; van Stekelenburg & Klandermans, 2007 for
overviews).

1
0162-895X © 2013 International Society of Political Psychology
Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, 9600 Garsington Road, Oxford, OX4 2DQ,
and PO Box 378 Carlton South, 3053 Victoria, Australia
2 Klandermans

Decades earlier, students of social movements had incorporated the concept of collective
identity in their theoretical frameworks (see Klandermans, Kriesi, & Tarrow, 1988). The weakness of
the social-movement literature on identity and contention, though, was that the discussion remained
predominantly theoretical and qualitative if at all empirical. Collective identity was declared to be
important but very little was said about why and how it was important. Few seemed to bother about
evidence. Indeed, in those days there was not even agreement on what exactly collective identity
would be like, let alone how it could be observed. Basic questions such as how collective identity is
formed, becomes salient, or politicized were neither phrased nor answered (see also Stryker, 2000).
Perhaps social-movement scholars didn’t bother too much because they tend to study contention
when it takes place and when collective identities are already formed and politicized.
Some 15 years ago, I began to explore the role of identity in political protest. Coming from a
social-psychological background, I was familiar with the work of Tajfel and his collaborators on
social identity (Tajfel & Turner, 1986). At first, I assumed that the social movement and social-
psychological literatures were addressing the same phenomenon and only employing different
labels, until I figured out that this wasn’t the case (Klandermans & de Weerd, 2000). Collective
identity in the social-movement literature is a group characteristic in the Durkheimian sense.
Someone who sets out to study that type of collective identity may look for such phenomena as the
group’s symbols, its rituals, beliefs, and the values its members share (cf. Taylor & Whittier, 1992).
Groups differ in terms of their collective identity. The difference may be qualitative, for example,
being an ethnic group rather than a gender group, or quantitative, that is a difference in the strength
of collective identity (Lofland, 1995). Social identity in the social-psychological literature is a
characteristic of a person. It is that part of a person’s self-image that is derived from the groups of
which he or she is a member. Social identity supposedly has cognitive, evaluative, and affective
components that are measured at the individual level (Drury, Reicher, & Clifford, 1999; Ellemers,
Spears, & Doosje, 1999; Simon, 1999). Individuals differ in terms of social identity, again both
qualitatively (the kind of groups with which they identify) and quantitatively (the strength of their
identification with those groups). I use the term “collective identity” to refer to an identity shared by
members of a group or category. Collective identity becomes politically relevant when people who
share a specific identity take part in political action on behalf of that collective.
Note that identity is not the only factor in the dynamics of protest participation. Elsewhere
(Klandermans, 2003), I have argued that identity is one of the three fundamental reasons why people
participate in political protest. The two other reasons are instrumentality and ideology. People may
want to take part in political protest because they want to change their circumstances (instrumental
motives), to act as a member of their group (identity motive), or to express their views and feelings
(ideological motive). Social movements may supply the opportunity to fulfil these demands and the
more they do, the more movement participation turns into a satisfying experience.
In this article, I elaborate on the relationship between identification and participation in political
protest and dwell on some unresolved issues in the protest and identity literature. First, I discuss the
concept of politicized collective identity, positioning it within the identity framework. Second, I deal
with sequence and causality. What comes first, identification or participation? As I will show, there is
evidence confirming both causal directions. Third, I deal with the fact that individuals have multiple
identities. I am male, I am a citizen of Amsterdam, I am a university professor, I am a supporter of a
green party, etc. These many identities do not necessarily work in the same direction. Once a protest
movement gains momentum, it tends to divide a society into movement allies and opponents. As a
consequence, individuals might come under cross pressure when groups they identify with end up at
opposite sides of the controversy. Most of the time, however, people are embedded in homogeneous
social networks and structures that encourage or discourage participation in collective action. Finally,
I will discuss the radicalization of politicized collective identity. Radicalization stems from a failure of
the state to absorb the demands of frustrated, marginalized groups. The group’s collective identity is
Presidential Address 3

politicized; the group has attempted to influence the state but failed. If then the group’s identity is
threatened by the majority or repressive authorities, the group radicalizes.

Politicized Collective Identity

Simon (1999) describes “identity” as a place in society. People occupy many different places.
They are student, unemployed, housewife, soccer player, politician, farmer, and so on. Some of those
places are exclusive, occupied only by a small number of people. The members of a soccer team are
an example. Others are inclusive, encompassing large numbers of people, such as “Europeans.”
Some places are mutually exclusive, such as male-female, or employed-unemployed; some are
nested, for example, Dutch versus European; and some are crosscutting, such as female and student.
Personal, social, and collective identity. The roles and positions a person occupies form his
personal identity. At the same time, every place a person occupies is shared with other people. I am
not the only professor of social psychology, nor the only Dutch or the only European. I share these
identities with other people—a fact that turns them into social and collective identities at the same
time. Social identity concerns the socially constructed cognitions of an individual about his mem-
bership in one or more groups. Collective identity concerns cognitions shared by members of a single
group about the group of which they are a member. Indeed, while social identity is a characteristic
of an individual and involves more than one group, collective identity is a characteristic of a group
and involves more than one individual. In sum, then, although they are related, social identity and
collective identity are concepts at different levels of analysis.1
Identity strength. Self-categorization theory proposes that people are more prepared to employ
a social category in their self-definition the more they identify with that category. Identification with
a group makes people more prepared to act as a member of that group (Turner, 1999). This assertion
refers to identity strength. In her review of social identity theory, Huddy (2001) observes that social
identity literature tends to neglect that real-world identities vary in strength. But, she argues,
identifying more or less strongly with a group may make a real difference especially in political
contexts. Moreover, she suggests, strong identities are less affected by context. Following this
reasoning, we may expect that strong identities make it more likely that people act on behalf of their
group.2
Group identification connects social and collective identity. It is akin to commitment to the
group. Group identification can be assessed in all kinds of ways, but any operationalization of group
identification will refer somehow to what it means to an individual to belong to the group in point and
will thus implicitly or explicitly refer to the pride and significance of being a member of the group,
to the symbols, the values, the fate shared by the group members. On the one hand, identification
with a group is a building block of someone’s social identity. As a rule, someone’s social identity is
construed from identification with various groups. On the other hand, no collective identity can be
construed without individual members who identify with the group involved.
Salient collective identity. Most of the time collective identities remain latent. Self-
categorization theory hypothesizes that depending on contextual circumstances, an individual may
act as a unique person, that is, display his personal identity or as a member of a specific group display
one of his collective identities (Turner, 1999; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987).
Self-categorization theory has focused on when and how a particular social identity becomes shared
and outlines this process through an analysis of social identity salience. When a social identity
becomes salient, there will be a shared collective identity and an increase in the strength of group
1
Burke (2004) and Stryker (2000) distinguish role identity in addition to person and social identity. “Person identity” refers
to the unique characteristics of a person, while “role identity” and “social identity” are shared with other people (role
occupants and group members).
2
Nation and religious community are examples of groups that tend to generate high levels of identification.
4 Klandermans

identification. The process that affects one’s own social identity will affect other’s social identity
salience leading to a shared collective identity because these processes are tied to the same reality.
This is not always a matter of free choice. Circumstances may force a collective identity into
awareness whether people like it or not, as the Yugoslavian and South African histories have
illustrated dramatically. But also in less extreme circumstances, collective identities can be imposed.
For example, Mannerini, Roccuto, Fedi, and Rovere (2009) show how plans to construct a high-
speed train in a Northern Italian valley turns being an inhabitant of that valley from a latent into a
highly salient collective identity that rapidly politicized.
Politicized collective identity. Simon and Klandermans (2001) hold that when people become
involved in political protest on behalf of a group, the collective identity of that group politicizes.
They define politicized collective identity as that form of collective identity that underlies group
members’ willingness to “engage, as a mindful and self-conscious collective (or as representatives
thereof), in. . . . a power struggle knowing that it is the wider, more inclusive societal context in
which this struggle takes place and needs to be orchestrated accordingly” (p. 323).
Collective identity politicizes when it becomes the focus of a struggle for power. Social groups are
often involved in power struggles in that they try to establish, change, or defend a power structure.
Politicized collective identity is not an on/off phenomenon. Instead, politicization of collective identity
and the underlying power struggle unfold as a sequence of politicizing events which gradually
transform the group’s relationship to its social environment. Typically, this process begins with the
awareness of shared grievances. Next, a political actor is blamed for the group’s predicament, and
claims for compensation are leveled against that actor. Unless appropriate compensation is granted, the
power struggle continues. If in the course of this struggle the group seeks to win the support of third
parties such as more powerful authorities (e.g., the national government) or the general public
collective, identity fully politicizes. The attempt to involve these parties in the power struggle
inevitably turns the issue into a matter of public or general interest. This final step also results in a
transformation of the group’s relationship to its social environment because involving a third party
implies recognition of society or the larger community (e.g., the city, region, country, or the European
Union) as a more inclusive ingroup. Politicized collective identity thus implies a cognitive restructur-
ing of the social environment into opponents and (potential) allies; this involves also strategic
reformulation of the conflict issue such that it appeals to potential allies as well.
Under those circumstances, dual identities—that is, simultaneous identification with a subor-
dinate and superordinate entity—develop. Indeed, politicized collective identities are dual identities.
Societal groups are embedded in the same superordinate political entity (e.g., the nation-state or
society at large); identification with this entity or its inhabitants comes into play as a consequence of
the process of politicization. Recent research by Simon (2011) suggests that holding a dual identity
directs people to moderate action. When identification with a superordinate entity declines, discon-
tent more likely results in radical action.

Causality

The Problem

Among the unresolved matters regarding identity and protest participation is the issue of
causality. Does identification precede participation, or does participation precede identification? To
determine this, it is important to distinguish between the processes of consensus mobilization and
action mobilization, two mechanisms that are crucial in the emergence of collective action. Con-
sensus mobilization concerns the dissemination of the social movement’s viewpoints; action
mobilization concerns the transformation of sympathizers into participants. Identification pro-
cesses play an important role in both mechanisms. The research in the next sections demonstrates
Presidential Address 5

that identification significantly influences processes of consensus mobilization and subsequently that
in the context of action, mobilization identification and participation mutually reinforce one another.

Identification and Consensus Mobilization

Various actors disseminate diverging views. Competing movement organizations, countermove-

ment organizations, and opponents all try to convince people. The claims that move social-movement
organizations forward are by definition controversial. As a consequence, various actors disseminate
diverging views (Goldstone, 2004; Klandermans, 1992). We assumed that people are more inclined
to subscribe to an actor’s viewpoint the more they identify with that actor. In other words, calls for
action are more likely to resonate among people who identify with the organizer of that action. In a
comparison of two competing demonstrations in Amsterdam, we collected evidence corroborating
this assertion.
On Saturday, 2 October 2004, more than 300,000 people gathered on two different squares of
Amsterdam in two separate demonstrations to protest against the government’s austerity plans. The
largest of the two demonstrations (250,000 participants) was organized by the labor unions, the smaller
(50,000 participants) by “Keer het Tij” (Turn the Tide, TtT), an anti-neoliberalism alliance. The two
demonstrations were three kilometers apart on two different locations in Amsterdam at exactly the
same time against the same budget cuts proposed by the government, but they were staged by different
organizers who emphasized different aspects of the policies proposed by the government.
It all began with the decline of the economy. The Dutch government had announced a compre-
hensive package of austerity measures (inter alia plans to reduce early retirement rights). The
government, employers’ organization, and the unions failed to reach an agreement regarding the
early retirement rights; the consultations broke down, and the government announced that it would
put its own plans through. The labor movement declared that in order to reclaim its position at the
negotiation table, it had no choice but to mobilize for collective action. TtT came from a different
angle. It opposed what it defined as a harsh right-wing climate in the country and antisocial
government policies.
At the heart of every protest are grievances. Although they do not provide sufficient reason for
people to take part in protest activities (e.g., McCarthy & Zald, 1976), grievances are the fuel of the
motivational engine. The more people feel that the interests of the group and/or the principles that the
group values are threatened, the stronger is their motivation to take part in protest to defend their
interests and principles. Interests and principles are linked to instrumental and ideological motiva-
tion. Instrumental motivation refers to participation as an instrument to improve the situation of one’s
group. Efficacy, that is the belief that collective action will be effective, is a crucial element of
instrumental motivation. Ideological motivation refers to participation in defense of principles and
values that have been violated. We assumed that instrumental motives were more likely to resonate
with campaigns that emphasize the violation of interests, while ideological motivation was more
likely to resonate with campaigns that emphasize the violation of principles. Significant in the
context of this article, I assume that appeals and motives resonate especially among people who
identify with organizations that stage demonstrations. The more individual members of a social-
movement organization identify with that organization, the more likely their motives resonate with
the frames provided by that organization. Group identification is an awareness of similarity, ingroup
identity, and shared fate with others who belong to the same category (Brewer & Silver, 2000). It has
pervasive effects on what people feel, think, and do (Terry & Hogg, 1996). The more people identify
with a social-movement organization, the more individual orientations, values, and beliefs become
congruent with those of the social-movement organization.
The labor unions were utterly frustrated by the fact that the government had announced it would
proceed without further consultation. As far as the unions were concerned, their goal was getting
6 Klandermans

access to the deliberations once more. Turn the Tide’s campaign was broader and more ideological
than that of the labor unions, mobilizing against neo-liberal policies and conservative politics. Thus
characterized, we presume that the campaign of the labor unions will be more about interests and that
of TtT about principles. Therefore, we expect the participants in the union demonstration—
especially those who identify with the organizers—to be more instrumentally motivated. Similarly,
we expect the participants in the TtT demonstration—especially those who identify with the
organizers—to be more ideologically motivated. This is exactly what we found in a survey among
participants in the two demonstrations.3
Figure 1 depicts the interaction of demonstration and identification. The interactions tell us that
the participants in the two demonstrations must be distinguished from those who identify strongly
with the organizations that staged the demonstration and those who identify weakly. For those who
identified strongly, the appeals issued by the organizers resonated with instrumental motives for the
union demonstration and ideological motives for the TtT demonstration. As a consequence, partici-
pants in the TtT demonstration who identified strongly with organizations staging that event were
highly ideologically motivated (Panel a), and in fact, among these participants, high levels of
identification reduced the instrumental motivation. In contrast, participants in the union demonstra-
tion who identified strongly with the union were highly instrumentally motivated (Panel b & c).
Indeed, identification with an organization makes people susceptible to the frames and inter-
pretations propagated by that organization. Thus, through its influence on the effects of persuasive
communication, identification indirectly impacts protest participation.

Identification and Action Mobilization

Next to influencing consensus mobilization, identification influences action mobilization. More

specifically, I refer to the determination and the timing of people’s decisions to take part in collective
action. The reasoning is the following. The stronger people’s identification with the organizers and
the other participants, the more they are determined to participate, the earlier they hear about the
event, and the earlier they decide to take part in the collective action. Indeed, this is what our research
among participants in street demonstrations reveals.
Between 2009 and 2012, we surveyed some 13,000 participants in 60 street demonstrations in
seven European countries (Belgium, Great Britain, Italy, the Netherlands, Spain, Sweden, and
Switzerland). Twenty-two were ritual parades such as May Day Parades, Climate Change Marches,
and Gay Prides; 38 were various protest demonstrations. Of those 38, 19 were protests against
austerity measures, and 19 were various other demonstrations. We asked participants how deter-
mined they were to take part in the demonstration (on a 5-point scale from “not very” to “very
much”) and when they had made a firm decision to participate (“the day of the demonstration”; “a
few days before the demonstration”; “a few weeks before the demonstration”; “over a month ago”).
One-third of the participants (32.5%) were firmly decided “over a month ago”; 29.9% were decided
“a few weeks before”; 30.9% decided “a few days before”; and 8.7% decided the very day of the
demonstration. As we expected, decision time is related to the firmness of the decision to take part
(Pearson correlation: .44): the more determined people were to take part in the demonstration, the
earlier they made a firm decision to participate.
Participants in ritual parades decided earlier to take part than participants in protests against
austerity measures, and participants in austerity protests decide earlier than participants in nonaus-
terity protests. This is partly due to the fact that participants in parades and antiausterity protests are
more embedded in organizers networks, while participants who are more embedded in organizers

3
We distributed questionnaires at the demonstration. In the Turn the Tide demonstration, 209 participants turned in the
questionnaires (42%), and 233 participants in the union demonstration turned them in (47%).
Presidential Address 7

0.4

0.3

0.2

Ideology
0.1

0
ident.low ident.high
-0.1

-0.2

-0.3

0.25
0.2
instrumentality-value

0.15
0.1
0.05
0
-0.05 ident.low ident.high

-0.1
-0.15
-0.2

0.5
0.4
Instrumentality-efficacy

0.3
0.2
0.1
0
ident.low ident.high
-0.1
-0.2
-0.3
-0.4

Turn The Tide

Union
Figure 1. Interaction of demonstration and identity.

networks identify more with the organizers and other participants. Note that participants who are
embedded in organizer networks identify with the organizers and other participants more than those
who are not embedded (Table 1). Note also that participants who are not embedded in organizer
networks identify more with other participants than with organizers.
8 Klandermans

Table 1. Decision Time and Embeddedness

Ritual Parade Austerity Nonausterity

Demonstrations Demonstrations
Decision time (mean) 3.1 2.8 2.6
Embedded in organizer 53% 65% 32%
networks
Embeddedness × Identification

Embedded Not Embedded

Identification with Identification with Identification with Identification with
Organizers Participants Organizers Participants
4.3 4.2 3.4 3.9

Table 2. Embeddedness, Identification, and Decision Time:

Standardized Regression Coefficients

Model 1 Model 2
Embeddedness .25 .18
Identification with organizers .12
Identification with participants .13
R2 .06 .10
N = 11,422

Relevant for our discussion, embeddedness and identification are predictors of decision time
(Table 2). People who are embedded in organizer networks decide earlier (Model 1), and embed-
dedness people who identify with organizers and other participants decide earlier as well (Model 2).
The interaction of embeddedness and identification did not contribute significantly to the equation.
In sum, identification with the organizers and the other participants influences people’s moti-
vation and determination to participate in collective action. It makes the framing of the organizers
resonate with people’s motivational configuration, makes them determined to participate, and makes
them decide to take part early.

From Identity to Participation to Identity

The basic assumption regarding identity and participation in political protest is fairly straight-
forward. The more someone identifies with a group, the higher the chances are that he or she will take
part in collective action on behalf of that group. The available evidence overwhelmingly supports this
assumption (de Weerd, 1999; de Weerd & Klandermans, 1999; Kelly, 1993; Kelly & Breinlinger,
1996; Kelly & Kelly, 1992; Klandermans & de Weerd, 2000; Klandermans, Roefs, & Olivier, 2001;
Simon, 2011; Simon et al., 1998; Stürmer, 2000). An example is Stürmer and Simon’s (2009) study
of student protest in Germany against plans to introduce tuition fees at German universities. These
authors report that the data replicate earlier findings with regard to the gay movement and fat-
acceptance movement. The study demonstrates that in addition to instrumental motives, identity
motives account for the students’ readiness to take part in protest actions. The authors interpret
identification with the movement as a sign of politicization and thus view their findings as a
confirmation of the role of the politicization of collective identity.
Coming from the opposite direction, Reicher and his colleagues (Drury, Reicher, & Scott, 1999;
Reicher, 1984, 1996) report on how collective action establishes and strengthens collective identity.
Presidential Address 9

These authors have demonstrated repeatedly that participation in collective action reinforces iden-
tification with the group. In a more recent article, Drury and Reicher (2009) elaborate on the
empowering experience collective action can be and the impact it has on identification and future
action participation. Drawing on field studies of crowd events, Drury and Reicher (2009) note how
police action creates a strong unified crowd out of an initially fragmented collectivity. Crowd
members then sense their strength which led them to challenge the police, thus developing events
from sporadic skirmishing into generalized conflict.
As most studies are correlational, they demonstrate that high levels of identification go together
with high levels of participation, but they do not allow for conclusions about causality. The few
studies that employed longitudinal or experimental designs are better suited to test causal relations,
but they are inconclusive. For instance, whereas Simon et al.’s (1998) findings suggest that collective
identity stimulates collective action participation, De Weerd and Klandermans (1999) did not find
such a direct link between identification and participation; they did find that group identification
affects action preparedness which in turn affects action participation. In any event, both studies
suggest that identification comes first and then participation.
A further complication relates to the fact that action participation presupposes some opportunity
to act. In the absence of such opportunities, it is impossible for identification to translate into action,
but identification might still reflect in high levels of action preparedness. In fact, Fishbein and Ajzen
(1975) have emphasized that cognitive variables such as beliefs and attitudes influence behavior
intentions and that the extent to which such intentions translate into actual behavior depends on
contextual factors such as freedom and opportunity to act. All this suggests a more complicated
model than the simple direct link between identification and collective action participation with
which I began this section. In a study among farmers in the Netherlands and Spain, we employed a
panel design which made it possible to test a more complex model.
In the winter of 1993/94, the winter of 1995, and the fall of 1995, we interviewed three times 167
Dutch farmers and 248 Galician4 farmers. During those two years, several agricultural measures to
be taken by the national governments or the European Union were imminent, varying in impact on
the agricultural sector. The study was designed to investigate farmers’ responses to these measures.
We had chosen the Netherlands and Galicia because they are similar as far as agricultural products
are concerned but opposites as far as agricultural development is concerned. In both countries, the
same kind of farmers were involved, mainly from dairy, arable and mixed farms, but on modern and
large farms in the Netherlands and old-fashioned and small farms in Galicia. Agriculture in the
Netherlands more than in Spain has been the subject of political intervention and regulation. In
response, farmers in the Netherlands early on established professional organizations that evolved into
a powerful agricultural lobby. Agriculture became one of the best organized sectors of the Dutch
economy. The establishment of the European Community has only added to the situation. Dutch and
European politics turned farmers in the Netherlands into a group with common interests and
opponents. Spain and certainly Galicia were less successful. Moreover, Spain entered the European
Community more than 20 years later than the Netherlands, and as a consequence, farmers in Spain
are much weaker organized than those in the Netherlands. Therefore, we expect a stronger sense of
collective identity among Dutch farmers than among Galician farmers. On the other hand, farmers’
organizations in Spain are much more politicized than their counterparts in the Netherlands. Spanish
farmers’ organizations are branches of political parties whereas Dutch farmers’ organizations are
professional organizations of which almost any farmer is a member. Thus, there are fewer members
of farmers’ organizations in Spain than in the Netherlands, but those in Spain who are members are
expected to be more politicized than their Dutch counterparts. This is what our data seem to tell us
(Table 3).

4
Galicia is the most northwestern province of Spain.
10 Klandermans

Table 3. Action Participation and Action Preparedness

Galicia Netherlands

1993/4 1995-1 1995-2 1993/94 1995-1 1995-2

1 2 2 1 2
Action participation 37.3% 2.8% 2.0% 86.8% 11.4% 17.4%2
Action preparedness3 2.59 2.41 2.50 2.84 2.75 2.80
Identification with farmers4 1.77 – 1.66 1.95 – 1.94
Participation in farmers organization5 .19 .17 .19 1.11 1.08 1.04
1
participation in the past; 2participation since the previous interview; 3on a scale from 1 to 5; 41 = no identification,
2 = identification; 50 = not member, 1 = member, 2 = active member

Table 4. The Causal Link Between Identity and Collective Action Participation: Cross Lagged Standardized Betas

a. Preparedness at T3 b. Preparedness at T1
Identification at T1 .20 Identification at T3 .06
Identity organization at T1 .22 Identity organization at T3 .18
R2 .11 R2 .04
c. Participation at T3 d. Participation at T1
Identification at T1 .03 Identification at T3 .17
Identity organization at T1 .27 Identity organization at T3 .39
R2 .08 .22

Dutch farmers identify more with other farmers (main effect country: F = 64.98, df = 1,
p < .0001) and are more often a member of a farmers’ organization (main effect country: F = 524.57,
df = 1, p < .0001) than Galician farmers. Altogether, these findings suggest that in the Netherlands,
identification with farmers is stronger than in Galicia. In line with our theoretical reasoning, action
preparedness (main effect country: F = 11.21, df = 1, p = .001) and participation (chi-squares
1993/4: 92.04, 1995-1: 12.44, 1995-2: 31.25, with df = 1 all p’s < .0001) are higher among Dutch
farmers. Overall, this pattern is fairly stable over time.
Although in line with our expectations, these results do not necessarily prove that it is identity
processes that generate the differences in protest preparedness and participation. In order to solve
that matter, we ran a series of cross-lagged regression analyses exploiting our longitudinal design.
Table 4 presents the cross-lagged beta’s resulting from these analyses. Preparedness to participate in
collective action and actual participation at T3 are regressed on identification and membership of
identity organizations at T1 (Panel a and c), indicating whether identification at T1 predicts pre-
paredness to participate and actual participation at T3. Subsequently, preparedness to participate and
actual participation at T1 are regressed at identification and membership of identity organizations at
T3 (Panel b and d), indicating whether action preparedness and actual participation at T1 predict
identification at T3.
The results show that identification fares better as a predictor of action preparedness (beta = .20)
than action preparedness as a predictor of identification (beta = .06). On the other hand, actual
participation fares better as a predictor of identification (.17) than identification as a predictor of
participation (.03). The same pattern was found for membership of identity organizations, although
it was less pronounced (.22 versus .18 and .27 versus .39). Indeed, the variance in action prepared-
ness explained by identification (.11) is almost three times higher than that of the variance in
identification explained action preparedness (.04). The variance of identification explained by actual
participation is almost three times higher than that of actual participation explained by identification
(.22 versus .08).
Presidential Address 11

Conclusion

Group identification appears to foster action participation, be it by impacting on the effects of

persuasive communication, on the firmness and timing of the decision to participate, on action
preparedness, or the translation of action preparedness into action participation. In its turn, action
participation strengthens group identification. Indeed, group identification, action preparedness, and
action participation seem to function as a recurrent, self-reinforcing mechanism. All this suggests
that high levels of identification generate the preparedness to participate in collective action on behalf
of the group. If such action is actually staged, chances are high that preparedness is transformed into
action. Participation in its turn reinforces strength of identification. This hinges at the possibility that
identity strength and movement participation are of a recursive nature. Social-movement literature
has forwarded this as a theoretical idea but thus far provided very little in terms of empirical support.
In fact, our findings suggest a synthesis of the two causal patterns we encountered in the literature—
from identity to participation and conversely from participation to identity. Note that the transfor-
mation of action preparedness into action participation functions as the link between the two identity
processes. In others words, only if the opportunity to act is offered so that participation can actually
take place does the strengthening of identity occurs.

Multiple Identities

The Problem

Individuals occupy many different places in society and thus have multiple identities. They share
these identities with other people so that each identity at the same time is a collective identity. As
some identities are associated with organizational memberships, someone’s social identity is also
reflected in his or her social embeddedness. Farmers are members of farmer’s organizations, workers
are members of labor unions or professional associations, migrants are members of cultural orga-
nizations, students of student associations, and so on. Some of those identities are more important for
an individual than others, and membership of identity organizations signifies the salience of an
identity for the individual. Not always are people’s identities in sync. Indeed, the groups people
identify with may end at opposite sides of a controversy. This happened, for instance, in the
Netherlands, to people who at the same time were against the deployment of cruise missiles and
supporters of the Christian Democratic Party that was in favor of deployment. Under the cross
pressure they experienced, they ended up choosing the side most of their friends chose (Oegema &
Klandermans, 1994). If most of their friends stood by their party, they refrained from supporting the
cruise missile protests; if their friends supported the protests, they supported it as well.
The literature on identification and participation tends to neglect the simultaneous existence
of various identities (but see Kurtz, 2002). As a consequence, little is known about the impact of
such multiple identities. In the following, I will first compare collective action driven by single and
multiple identities, and then I will concentrate on a specific type of multiple identity, namely dual
identity. Together, the evidence reveals that the working of identity, especially multiple identity, is
complex and far from clear.

Single and Multiple Identity Protest

The fact that individuals have multiple identities raises the question of why specific identities
become the rallying points of protest movements. Shared grievances is part of the answer. The
awareness that grievances are shared makes a collective identity salient. Attempts to mobilize those
who share the same identity in order to redress the grievances further increase the salience of the
12 Klandermans

Table 5. Social Embeddedness and Protest Participation by Immigrants

Past Participation Pearson r Pearson r Pearson r

in Protest Unfairness × Efficacy × Political Discussion ×
Participation Participation Efficacy
No membership 25% .04ns .09ns X
Membership cso’s 50% .18 .28 .46
Membership identity org. 60% .33 .40 .66
n = 306

collective identity. Protest movements in reaction to a sudden imposed grievance are illustrations in
point (Walsh, 1981). Grievances may concern single identities or multiple identities. Single-identity
protests are reactions to identity politics that (are deemed to) disadvantage a specific group—be it
women, migrants, farmers, or religious or ethnic communities. Multi-identity protests react to
general political issues that affect a broad range of citizens—for instance, environmental issues,
peace and war, or global justice. Identity and issues politics are contextual matters that make
different identities salient. People are becoming aware or made aware of the fate they share.
Movement organizations and organizers try to mobilize the people affected. In doing so, they make
people aware of the identity they share. As identity and issue politics vary between nations and over
time, we will find different protest events in different countries and at different points in time.
However, if identity or issue politics coincide, we may find comparable protest events in different
countries (cf. Walgrave & Rucht, 2010).
Protest events do not originate in a vacuum. Organizers must first build mobilizing structures
and then employ these structures to reach potential participants (Boekkooi, Klandermans, & van
Stekelenburg, 2011). These authors demonstrate that the composition of the mobilizing structure
influences the composition of the demonstrating crowd. Organized identities more likely than
unorganized identities become the rallying point of political protest if those identities are threatened
one way or another. Farmers in our protest study who were members of farmers’ organizations were
more likely to take part in a farmers’ protest than farmers who were not members of such organi-
zations. Citizens who are members of multiple organizations are more likely to participate in
multiple identity protests (Bennett, Givens, & Breunig, 2010). These authors coin the phrase
“complex political identification” for individuals involved in several social and political organiza-
tions at the same time. Obviously, mobilizing structures may or may not fit with the social embed-
dedness of would-be participants. I started this section by linking identity with social embeddedness.
Let me give two illustrations of how this link influences protest participation—one example of a
single-identity protest and one of a multi-identity protest.
The first example concerns participation in political protest by Turkish immigrants in New York
and Turkish and Moroccan immigrants in the Netherlands (Klandermans et al., 2008).5 Roughly
two-thirds (62%) of our respondents were members of civil society organizations such as political
parties, religious organizations, cultural organizations, women’s organizations, educational organi-
zations, trade unions, residents committees, neighborhood committees, sports organizations, and
youth organizations. In follow-up questions, we assessed whether it concerned organizations exclu-
sively for immigrants or organizations for citizens in general; 21% of our respondents were members
of exclusively ethnic organizations, that is to say embedded in identity organizations. Table 5
presents figures that demonstrate the impact of general social embeddedness and embeddedness in
identity organizations on protest participation.

5
Using snowball sampling, we approached 126 Turkish immigrants and 80 Moroccans in the Netherlands and 100 Turkish
immigrants in New York.
Presidential Address 13

Table 6. Mobilization Deficit: Over- Versus Underrepresentation

U.K. Spain Italy Netherlands Switzerland Belgium Germany Total

Left +24 +23 +43 +22 +34 +30 +25 +30
Centre +2 −6 −13 +12 −21 +7 +9 −2
Right −27 −17 −28 −34 −13 −37 −34 −28
N 1116 448 1002 528 629 503 769 5661

Clearly, protest participation increases with social embeddedness (Column 1). Immigrants who
are involved in civil-society organizations did participate more frequently in political protest than
those who are less involved in politics—the more so if it concerns identity organizations. The
remaining three columns suggest reasons why this is the case. In the first place, discontent is more
likely to be translated into protest participation among those who are organizationally embedded,
especially in migrant organizations (Column 2). Secondly, perceived unfairness is correlated stronger
with collective action participation among people who are members of civil-society organizations,
again especially migrant organizations. The reason why this is the case is suggested by the figures in
Columns 3 and 4. First, people who are embedded in civil-society organizations, especially exclu-
sively migrant organization, feel more efficacious (Column 3). This is presumably so because in
those organizations, they discuss and learn about politics (Column 4).
The second example comes from a study of the world-wide demonstrations against the war in
Iraq on February 15, 2003 (Walgrave & Rucht, 2010).6 The protests against the war in Iraq are typical
examples of multi-identity protests in response to issue politics. A broad scala of identities all leaning
to the left of the political spectrum was triggered by the issue. Therefore, we expected that partici-
pants would share a leftist political identity and be embedded in networks of the left, and this is what
we found (Table 6). A comparison of left-right self-placement of participants in the demonstration
with that of the general population revealed that large proportions of the left of the political spectrum
were overrepresented (+). On the other hand, the right-wing of the political landscape was clearly
underrepresented (−). In other words, social embeddedness in organizations and networks of the left,
including a leftist political identity, made protest participation more likely; embeddedness in net-
works of the right and a rightist political identity made participation in the demonstration less likely.

Dual Identity

A special example of the dynamics of multiple identities are dual identities, that is, the going
together of superordinate identities such as national identity and subgroup identities such as ethnic-
ity, class, religion, or gender. Any nation is a conglomerate of many different population groups. The
political common sense seems to be that identification with a subgroup and identification with the
nation are mutually exclusive. Be this as it may, throughout history, governments have tried to foster
national identification and to suppress expressions of subgroup identification. For instance, in
Franco’s Spain, every sign of regional identification was brutally suppressed (Johnston, 1991).
Protest groups are easily blamed for being disloyal, unpatriotic, and forsaking the national identity,
as the fate of many protest movements on the African continent and the peace movements in the
United States and Europe during the Cold War illustrated. This is true even more so for immigrants.
Immigrants occupy a delicate position in their “host” societies. They are expected to assimilate into
the host culture, and failure to do so is considered a sign of a lack of loyalty. Immigrants’ loyalty to
the country of residence is placed under even more doubt if they engage in protest. However, history

6
Walgrave and Rucht (2010) surveyed 5,661 demonstrators in eight countries (United States, United Kingdom, Italy, Spain,
Netherlands, Germany, Belgium, and Switzerland).
14 Klandermans

has demonstrated time and again that the suppression of subgroup identity reinforces rather than
weakens subgroup identification (Zegeye, Liebenberg, & Houston, n.d.). This suggests that national
identity and subgroup identities are not necessarily mutually exclusive, but rather they are made
mutually exclusive through identity politics which suppresses subgroup identities in favor of national
identity.
Political protest presupposes some subgroup identity. After all, some subgroup in society is
aggrieved. Therefore, the interplay of national identity and subgroup identity is an important aspect
of any theory of collective identity and political protest. The question I want to address here is how
national identity and subgroup identity are related to political protest. We know that subgroup
identification fosters participation in political protest on behalf of the subgroup but doesn’t that
weaken national identification? Alternatively, does a strong national identity make it unlikely for
people to be prepared to participate in collective action on behalf of some subgroup? In trying to
answer these questions, I will draw on recent social psychological research on dual identity (Brown,
2000; González & Brown, 2003; Huo et al., 1996; Klandermans et al., 2001, 2008; Simon, 2011).
This research starts from the distinction between subordinate and superordinate identities. The
available evidence suggests that subordinate and superordinate identities are far from mutually
exclusive. To the contrary, a combination of sub- and superordinate identity seems to contribute to
the stability of a political system. Indeed, intergroup hostility seems to decline if groups are allowed
to nourish both sub- and superordinate identities. González and Brown (2003; see also Brown, 2000)
have shown that a dual identity, that is, a strong superordinate identity in combination with a strong
subgroup identity, reduces the likelihood of conflicts between subgroups. Huo et al. (1996) demon-
strate that a sufficiently high level of superordinate identification makes it possible for authorities to
maintain cohesion within an ethnic diverse community. But studies such as those of González and
Brown (2003) emphasize that superordinate identity only has this effect if people can maintain their
subgroup identity as well. These studies show that enforcing people to forsake their subgroup
identity in favor of a superordinate identity backfires and generates separation rather than integration.
It is not always easy to decide what social psychological mechanisms are at work here. I suggest
as a possible mechanism that a strong national identity makes it possible for people to look at
subgroups in a different way and to accept that subgroups in a society are treated differentially.
Indeed, Americans who identify with “Americans” in addition to being “White” take a more positive
stand towards affirmative action (Smith & Tyler, 1996). And for East Germans, the awareness that
“we are all Germans” made it easier to accept the differences between East and West Germany
(Mummendey, Klink, Mielke, Wenzel, & Blanz, 1999). Tyler, Boeckmann, Smith, and Huo (1997)
argue that this is so because superordinate identity changes the relationship between subgroups and
authorities. These authors hold that the absence of superordinate identification makes that relation-
ship solely instrumental—what are authorities doing for my group. Superordinate identity makes it
possible for people to accept disadvantages done to their subgroup in the interest of the larger
community. People trust authorities to make sure that next time their group will benefit. This implies
that superordinate identity and trust in authorities are intimately related. But superordinate identity
should not undermine subgroup identity, dual-identity research seems to warn. All this suggests that
there is no reason to assume that national identity and subgroup identities are mutually exclusive.
What is more, dual identities seem to be beneficial to a political system. Let me conclude this section
with two examples from our own research—a study among South African citizens7 and a study
among migrants in the Netherlands and New York mentioned previously.
South Africa’s population is composed of many different population groups. Since the disman-
tling of apartheid, South Africa attempts to build a rainbow nation with which people of all walks of

7
Between 1994 and 2000, we interviewed yearly separate samples of over 2,100 South African citizens. The interviews
contained questions about political attitudes and behavior.
Presidential Address 15

Table 7. National Identity and Subgroup Identity: Percentages

South Africa 2000 Migrants NL/NY

No strong subgroup identity/no strong national identity 18 27
Strong subgroup identity/no strong national identity 40 33
No strong subgroup identity/strong national identity 6 16
Strong subgroup identity/strong national identity 36 23

Table 8. Dual Identity and Politics: South Africa

Trust in Prepared to Participate

Government** in Peaceful Action**
No strong subgroup identity/no strong national identity 2.7 2.4
Strong subgroup identity/no strong national identity 2.7 2.6
No strong subgroup identity/strong national identity 2.8 2.6
Strong subgroup identity/strong national identity 3.2 2.8
*F/df 3 = 79.88, p < .001. **F/df 3 = 39.81, p < .001.

life can identify. This raises the question of whether subgroup identities compete with national
identity. Between these two types of identities, individual citizens can take four different positions:
they can develop a strong subgroup identity but no national identity; a strong national identity and
no subgroup identity; or a dual identity. Of course, they can also develop neither identity. Immigrants
are in a comparable position: they might identify solely with their country of origin or with the
country of immigration; they might identify with both or with neither country (Table 7). Berry (1984)
labeled these four forms of cultural adaptation separation, assimilation, integration, and marginal-
ization, respectively. In both studies, we assessed whether our respondents identified with a sub-
and/or superordinated entity—subgroup and nation in South Africa and home country and country of
immigration for the immigrants.8 A glance at Table 7 suggests that in both cases the two identities are
far from mutually exclusive. Indeed, subordinated and supraordinated identification correlated at .23.
One-third of the South African citizens and about a quarter of the immigrants we interviewed
displayed dual identities.
But what about political attitudes and behavior? Did subgroup identity and national identity
interact in their influence on political attitudes and behaviour, and if so, did the two reinforce or
undermine one another? In order to answer this question in the South African context, we compared
the four groups on two variables regarding people’s relation to politics, namely trust in government
and people’s preparedness to take part in peaceful collective action. The results in Table 8 confirm
our expectations. Trust in government and action preparedness were the highest among those
respondents who displayed a dual identity. Interestingly, the figures for trust in government show that
those with a dual identity stand out. This suggests that at least as far as trust in government is
concerned, it is the combination of subgroup and national identity that made the difference. Simi-
larly, note that action preparedness was the highest among those with a dual identity.
Similar results were found in the migrants’ study. We assumed that dual identity would
be associated with more satisfaction with one’s situation. In order to test this assumption, we
compared those who displayed a dual identity with those who did not in terms of distributive and

8
In the migrants study, we asked our respondents whether they “feel connected” and “liked to be seen as” [American,
Netherlander, Turk, Moroccan]. This was turned into identity scales from 1 (“no identification”) to 5 (“strong identifica-
tion”). In the South African study, we asked “how close one feels to [group]” to be answered on a 5-point scale from 1 (“not
close”) to 5 “extremely close”) and “Being a South African, is it an important part of how I see myself” (do not agree, agree,
strongly agree).
16 Klandermans

procedural-fairness judgments. Respondents who displayed a dual identity judged the way the
government of the country of residence treated them—individually and as a group—as fairer than
those who did not display a dual identity (eta’s ranging between .13 and .20; p’s between .022 and
.001). As hypothesized, dual identity comes with more satisfaction about the way government treats
people. As a consequence, identity, especially ethnic identity, was negatively related to protest
participation, albeit modestly (Pearson r = −.13). This negative correlation implies that respondents
who identified more with their ethnic group participated less in collective action. We also checked
whether immigrants with a dual identity were more or less likely to take part in collective action, but
no such relationship was found. However, dual identity moderates the relationship between fairness
judgments and action participation: The positive relationship between perceived unfairness and
participation is only found among immigrants who display a dual identity (Pearson r = .23, p < .01
among those who display a dual identity versus .09, n.s. among those who do not). In other words,
immigrants with a dual identity are more satisfied with their situation, but if they feel treated unfairly,
they were more likely to engage in collective action than people who felt treated unfairly but do not
display a dual identity. No such effects were found for national or ethnic identity.
Thus, national identity and subgroup identity are not mutually exclusive. In fact, we found a
moderate, positive correlation which argues against all those who assume that protestors are less
loyal to the larger community. Dual identity—a combination of strong national and strong subgroup
identity—fostered trust in government and preparedness to take part in peaceful action in South
Africa and among immigrants satisfaction with their situation and action preparedness among those
who were dissatisfied.

Conclusion

Individuals have multiple identities, identities that reflect in how they are socially embedded.
Identity or issue politics make identities salient. Depending on the interaction with opponents and
allies, a collective identity politicizes. Politicization of collective identity, supposedly, intensifies the
impact of collective identity on action preparedness. Politicization of collective identity implies that
the involvement of the wider social and political environment is sought. Such efforts, I hold, require
some commitment to or identification with that environment. Why bother to change a nation if one
feels no commitment to it whatsoever? Simon (2010) thus argues that politicized collective identity
is dual identity. Indeed, politicized collective identity presupposes some level of identification with
a superordinate entity, that is, the nation. This is why South Africans were more likely to participate
in collective action if they display a dual identity and why immigrants if they are aggrieved have
more frequently participated in collective action when they have a dual identity.
Sidanius, Feshbach, Levin, and Pratto (1997) and Staerklé, Sidanius, Green, and Molina (2005)
hold that among majority groups in a society, subordinated and superordinated identification are
more likely to correlate positively than among minority groups. In other words, dual identification is
more likely among majority groups than among minority groups. This is corroborated by data from
studies among majority and minority students in the United States and Israel. We tried to replicate
this in our South Africa study with mixed results. A complicating matter in the South African context
is that it is not always clear what is the majority and what the minority. Be that as it may, taking the
Blacks as the majority and the Whites, coloureds, and Asians in that order as minorities, indeed less
Whites than Black South Africans displayed a dual identity (28% versus 38%); and coloureds and
Asians even less so (25% and 15%). However, correlations between sub- and supraordinated
identification did not corroborate Sidanius et al.’s (1997) proposition. Among both Blacks and
Whites, we found positive correlations of .20; among coloureds and Asians of .30 and .12, respec-
tively. But, I was interested in collective action participation among people who do display a dual
identity, be it majority or minority. Of those people, I would expect that they are more likely to be
Presidential Address 17

satisfied and less likely to take part in collective action. And, I maintain that if they are dissatisfied,
both majority and minority groups are more likely to take part in collective action if they display a
dual identity.
In addition to its presence as dual identity, politicized collective identity is also manifest in
membership of identity organizations. Remember that in a previous section I reported that for
participants in street demonstrations, farmers as well as migrants had higher levels of protest
participation among members of identity organizations than among nonmembers. In that section, it
was also reported that among migrants, members of identity organizations more frequently discussed
politics than other migrants.

Radicalization

The Problem

Simon and his team (cf. Simon, 2010) conceptualize radicalization as the process that turns a
dual identity into a separatist identity—a more exclusive form of collective identification lacking the
pacifying effect of superordinate identification. Radicalization—often driven by a failure of the state
to absorb the demands of frustrated, marginalized groups—facilitates disruptive forms of political
engagement including violent action. In this final section, I show how real-world events radicalize
people depending on processes of identification.

How World Events Radicalize Identities

Over the last few decades, the world has witnessed many examples of radical identity politics.
Extreme-right groups and parties openly express and disseminate xenophobic ideas in Western
European countries such as Austria, Belgium, Denmark, France, Italy, Norway, Switzerland, and the
Netherlands. In the United States, Christian fundamentalists, militias, and extreme-right groups
employ radical tactics. At the other end of the ideological spectrum, far-left groups like the antiglo-
balization movement, the “Black Block,” or radical environmentalists and animal-rights groups
engaged in radical actions. But undoubtedly, from a Western perspective, all this is overshadowed by
Muslim activism, be it the attack of the World Trade Center in New York, the trains in Madrid, the
subway in London, or the murder of Theo van Gogh in the Netherlands.
People do not radicalize on their own but as part of a group in which a collective identity is
embedded. Some members of the group take an activist route to promote or prevent change. Their
interactions with their opponents intensify, while their ideas and beliefs sharpen. Indeed, radicaliza-
tion is conceived of as a process of identity transformation. Nobody is born radical; radical political
actions result from processes of radicalization. Nested in the process of radicalization are two related
processes, namely politicization and polarization. Politicization I discussed already, but a few words
on polarization. I conceive of polarization as an instance of movement-countermovement dynamics
in which the two antipodes tend to “keep each other alive” (Meyer & Staggenborg, 1996). The two
sides reinforce each other, identifying themselves in opposition to each other and regarding the
others as the main target of their collective actions. When groups polarize, a sharp distinction
between “us” and “them” evolves. Both groups assert that what “we” stand for is threatened by
“them,” tribute is paid to the ingroup’s symbols and values, and the out-group is derogated.
Radicalization and intergroup conflict are interpreted as collective processes.
As a rule, the social-movement sector is internally fragmented along radical and moderate lines
over what constitutes appropriate means and desirable ends (Tarrow, 1998). Such fragmentation
might lead groups to break away seeking more radical goals and actions. By breaking away,
radicalizing group members turn their back to the society at large and to their fellow-activists. This
18 Klandermans

implies a “double marginalization,” both from society and the movement (della Porta, 1995, p. 107).
Such double marginalization results in material, social, and psychological isolation explaining the
group’s deviation from the “normal” perception of reality. Because of their isolation, group members
come into significant contact only with others who share their worldview and emotions. No deviance
of the group norm is accepted, and the degree of freedom declines to nearly zero.
Identity processes play a crucial role in polarization. Research demonstrates that threats from
countermovements can shape a movement’s collective identity (Einwohner, 2002). First of all, the
presence of powerful opponents makes identities more salient for activists (van Dyke, 2003).
Secondly, polarization implies a split in terms of friends and foes. Polarization, finally, also induces
a strategic reformulation of “who we are.”
Groups may also radicalize in processes of mutual confrontation. If in the heat of the confron-
tation people feel that their identity is threatened, this spurs identification processes that in turn
reinforce the confrontation. Social categorization makes people more aware of their group member-
ship which strengthens their identification. Due to processes of self-enhancement, group members
attempt to “repair” their threatened self-esteem through ingroup favouritism and out-group deroga-
tion. By portraying “us” as good and “them” as evil, the debate radicalizes; the groups drift apart and
polarize. In case of a protracted intergroup conflict, each new incident will start off the whole
process—threatened identity, attempts to repair, in-group favouritism, and out-group derogation—
over and again. Hence, from incident to incident, group identification becomes stronger and ideas
and feelings become more radical. This cyclical mechanism radicalizes the debate and widens the
cleavage between groups.
These days processes of radicalization can be observed at the Internet (Twitter, blogs, text
messages, Facebook). We studied the exchanges at two web forums in the Netherlands—a Moroccan
(Marokko.NL) and an ethnic Dutch (NL.politiek). At both networks, participants react to “real-world
events” and to each other.9 Figure 2 relates the amount of participation on the forums to identity
relevant events between October 2003 and April 2006, the period under study. The amount of
participation in both forums is expressed as the number of words in the postings about immigration
and integration. Obviously, the online discussion shows a strong response to the three major events
during this period.
After an event, the degree of participation on both sites increases strongly: by 55% for the ethnic
Dutch, and as much as 98% for the Moroccans. Also, the number of participants boosts, 43%
increase for NL.politiek, but especially for Marokko.NL (86%). In the four weeks after an event, the
number of negative expressions increases on average with 27%. On Marokko.NL, debaters mention
ethnic Dutch as out-group 33% more often, and they also refer to Muslims as an ingroup more often
(24%). On NL.politik, debaters reference to Muslims as an out-group increases with 19%, whereas
reference to ethnic Dutch as an ingroup decreases with 31%. In times of an intergroup conflict,
positive ingroup and negative out-group identification differ for dominant and subordinate groups.
Whereas group members of the dominant ethnic Dutch group display more out-group derogation
than ingroup favoritism, members of the subordinate Muslim group display both more ingroup
favouritism and out-group derogation.
Emotions are straightforwardly measured by counting how frequent emotion words (including
synonyms and variants) are mentioned. Obviously, severe intergroup incidents result in increased
mentioning of fear, anger, and hate on the web forums. Normally, in periods without incidents, the
three emotions are about equally often mentioned on both forums, although fear prevails more on the
forum of the ethnic Dutch. After an incident, the number of emotions mentioned raises much stronger
on Marokko.NL (24%) than on NL.politiek (9%). On Marokko.NL fear raises only slightly (9%)

9
From October 2003 to April 2006, we collected 40,051 postings of NL.politiek and 17,768 of Marocco.NL. At NL.politiek,
we counted more than 40,000 participants, at Marokko.NL over 89,000.
Presidential Address 19

300000
Murder Theo NlPolitiek
MarokkoNl
van Gogh
250000
Danish
Attack Attack cartoons
200000 Madrid London
Number of Words

Memorial
van Gogh
150000

100000

50000

200310 200312 200402 200404 200406 200408 200410 200412 200502 200504 200506 200508 200510 200512 200602 200604

Year-Month

Figure 2. Attention for immigration and integration issues on two opposing web forums (number of words).

after incidents, but anger (26%) and especially hate (39%) increase strongly. Fear does not only
prevail more on NL.politiek, it also increases more after incidents (17%) than anger (1%) and hate
(11%).

Conclusion

Obviously, devastating events such as the attacks in Madrid and London or the murder of Theo
van Gogh are the talk of the town as well as in the “virtual cafés.” After an incident, debaters clearly
feel the urge to express their view: participation on the web forums increases and for the Moroccan
forum even doubles. People “talk” much more in terms of the conflicting groups—Muslims, Jews,
Moroccans, ethnic Dutch, and son on—which enhances their inclination to define themselves as a
member of one of these groups—“I as Muslim”; “I as Dutch.” Incidents bring group membership
forcefully to the psychological foreground, such that Dutch Muslims and ethnic Dutch feel their
group identity threatened. Consequently, group members reveal the urge to defend or “repair” their
threatened group identity, and the debates boil down to a few stereotypical characteristics fueled by
strong emotional loadings to differentiate between the liked ingroup and the hated out-group. The
result: further polarization and radicalization.

Coda

Identity processes occupy a central place in psychological and sociological theory and research
of political-protest participation either as antecedents, mediators, moderators, or consequences. As
identities are always shared with other people, they are individual and collective at the same time. Yet
without identification, collective identities would be irrelevant. There is no politicization of collec-
tive identity without identification. However, people have many different identities at the same time.
20 Klandermans

That raises the question of why a specific identity becomes the focal identity in a specific protest
event. Shared grievances are at the root of political protest and thus of the politicization of collective
identity. Organizers interpret grievances, translate them into claims addressed at authorities, create
mobilizing structures, and seize opportunities to protest. Indeed, collective identity and group
identification are factors needed for protest to emerge, but appealing opportunities to participate and
effective mobilization are needed as well. Grievance interpretation, identity formation, and politici-
zation are taking place within social networks. That is why social embeddedness in such networks
is crucial in processes of politicization and mobilization. Politicization of collective identity takes
place when grievances are turned into claims and citizens begin to campaign and mobilize to win
support for their cause. In the process of politicization, the social environment changes into allies and
opponents.
A special form of multiple identities which is relevant in this context is dual identity. Politi-
cized identity is a form of dual identity—identification with a subgroup (ethnicity, class, gender,
religion, and so on) combined with national identification. Citizens who display a dual identity are
more satisfied with their situation, but if they are dissatisfied, they are more likely to engage in
collective action. Politicization can lead to radicalization, when authorities fail to respond to the
grievances and claims forwarded or choose to repress or because the relation between opponents
becomes more tense. When under such circumstances the moderating force of superordinate iden-
tification fades, it is just the subordinate identification that is left—a force that easily leads to
separation and radicalization.

ACKNOWLEDGMENTS

This article is a revised version of the Presidential Address read at the 35th Annual Meeting of
the International Society of Political Psychology, July 8, 2012, Chicago, IL. Correspondence con-
cerning this article should be sent to P.G. Klandermans, VU University, Dept. of Sociology, de
Boelelaan 1081, 1081HV, Amsterdam, The Netherlands. E-mail: p.g.klandermans@vu.nl

REFERENCES

Azzi, A. E., Chryssochoou, X., Klandermans, B., & Simon, B. (2010). Identity and participation in culturally diverse
societies. Oxford, UK: Blackwell Wiley.
Berry, J. W. (1984). Cultural relations in plural societies: Alternatives to segregation and their sociopsychological implica-
tions. In N. Miller & M. Brewer (Eds.), Groups in Contact (pp. 11–27). San Diego, CA: Academic Press.
Bennett, W. L., Givens, T. E., & Breunig, C. (2010). Crossing political divides: Communication, political identification and
protest organization. In S. Walgrave & D. Rucht (Eds.), The world says no to war. Demonstrations against the war on
Iraq (pp. 215–238). Minneapolis: University of Minnesota Press.
Boekkooi, M., Klandermans, B., & van Stekelenburg, J. (2011). Quarrelling and protesting. How organizers shape a
demonstration. Mobilization, 16, 223–242.
Brewer, M. B., & Silver, M. (2000). Group distinctiveness, social identity and collective mobilization. In S. Stryker, T. J.
Owens, & R. W. White (Eds.), Self, identity, and social movements (pp. 153–171). Minneapolis: University of Minnesota
Press.
Brown, R. (2000). Social Identity Theory: Past achievements, current problems and future challenges. European Journal of
Social Psychology, 30, 745–778.
Burke, P. J. (2004). Identities and social structures: The 2003 Cooley-Mead Award address. Social Psychology Quarterly, 67,
5–15.
Della Porta, D. (1995). Social movements, political violence and the state. Cambridge, UK: Cambridge University Press.
De Weerd, M., & Klandermans, B. (1999). Group identification and social protest: Farmer’s protest in the Netherlands.
European Journal of Social Psychology, 29, 1073–1095.
Drury, J., & Reicher, S. (2009). Collective psychological empowerment as a model of social change: Researching crowds and
power. Journal of Social Issues, 65, 707–725.
Presidential Address 21

Drury, J., Reicher, S., & Clifford, S. (1999). Shifting boundaries of collective identity: Intergroup context and social category
change in an anti-roads protest. British Psychological Society Social Psychology Section Conference. University of
Surrey, July.
Einwohner, R. L. (2002). Motivational framing and efficacy maintenance: Animal rights activists’ use of four fortifying
strategies. Sociological Quarterly, 43, 509–526.
Ellemers, N., Spears, R., & Doosje, B. (1999). Social identity. Oxford, UK: Blackwell.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behaviour. Reading, MA: Addison-Wesley.
Goldstone, J. A. (2004). More social movements or fewer? Beyond political opportunity structures to relational fields. Theory
and Society, 33, 333–365.
González, R., & Brown, R. (2003). Generalization of positive attitude as a function of subgroup and superordinate group
identification in intergroup contact. European Journal of Social Psychology, 33, 195–214.
Huo, Y. J., Smith, H., Tyler, T. R., & Lind, E. A. (1996). Superordinate identification, subgroup identification, and justice
concerns: Is separatism the problem; is assimilation the answer? Psychological Science, 7, 40–45.
Huddy, L. (2001). From social to political identity: A critical examination of Social Identity Theory. Political Psychology, 22,
127–156.
Iyer, A., & van Zomeren, M. (2010). Social and psychological dynamics of collective action. From theory and research to
policy and practice. Journal of Social Issues, 66(1).
Johnston, H. (1991). Tales of nationalism: Catalonia, 1939–1979. New Brunswick, NJ: Rutgers University Press.
Kelly, C. (1993). Group identification, intergroup perceptions and collective action. European Review of Social Psychology,
4, 59–83.
Kelly, C., & Breinlinger, S. (1996). The social psychology of collective action. Basingstoke, UK: Taylor and Francis.
Kelly, J., & Kelly, C. (1992). Industrial action. In J. F. Hartley & G. M. Stephenson (Eds.), Employment relations. The
psychology of influence and control at work (pp. 246–270). Oxford, UK: Blackwell.
Klandermans, B. (1992). The social construction of protest and multi-organizational fields. In A. Morris & C. Mueller (Eds.),
Frontiers in social movement theory (pp. 77–103). New Haven, CT: Yale University Press.
Klandermans, B. (2003). Collective political action. In D. O. Sears, L. Huddy, & R. Jervis (Eds.), Oxford handbook of political
psychology (pp. 670–709). Oxford, UK: Oxford University Press.
Klandermans, B., & de Weerd, M. (2000). Group identification and political protest. In S. Stryker, T. Owens, & R. W. White
(Eds.), Self, identity, and social movements (pp. 68–92). Minneapolis: University of Minnesota Press.
Klandermans, B., Kriesi, H. P., & Tarrow, S. (1988). From structure to action: Comparing movement participation across
cultures. Greenwich, CT: JAI Press.
Klandermans, B., Roefs, M., & Olivier, J. (2001). The state of the people: Citizens, civil society and governance in South
Africa, 1994–2000. Pretoria, South Africa: Human Science Research Council.
Klandermans, B., van de Toorn, J., & van Stekelenburg, J. (2008). Embeddedness and identity: How immigrants turn
grievances into action. American Sociological Review, 73, 992–1012.
Kurtz, S. (2002). All kinds of justice: Labor and identity politics. Minneapolis: University of Minnesota Press.
Lofland, J. M. (1995). Charting degrees of movement culture: Tasks of the cultural cartographer. In H. Johnston & B.
Klandermans (Eds.), Social movement and culture (pp. 188–216). Minneapolis: University of Minnesota Press.
Mannarini, T., Roccuto, M., Fedi, A., & Rovere, A. (2009). Six factors fostering protest: Predicting participation in locally
unwanted land uses movements. Political Psychology, 30, 895–920.
McCarthy, J. D., & Zald, M. N. (1976). Resource mobilization and social movements: A partial theory. American Journal of
Sociology, 82, 1212–1241.
Meyer, D. S., & Staggenborg, S. (1996). Movements, countermovements, and the structure of political opportunity. American
Journal of Sociology, 101, 1628–1661.
Mummendey, A., Klink, A., Mielke, R., Wenzel, M., & Blanz, M. (1999). Socio-structural characteristics of intergroup
relations and identity management strategies: Results from a field study in East Germany. European Journal of Social
Psychology, 29, 259–286.
Oegema, D., & Klandermans, B. (1994). Non-conversion and erosion: The unwanted effects of action mobilization. American
Sociological Review, 59, 703–722.
Reicher, S. D. (1984). The St. Paul’s riot: An explanation of the limits of crowd action in terms of a social identity model.
European Journal of Social Psychology, 14, 1–21.
Reicher, S. D. (1996). The Battle of Westminster: Developing the social identity model of crowd behaviour in order to explain
the initiation and development of collective conflict. European Journal of Social Psychology, 26, 115–134.
Sidanius, J., Feshbach, S., Levin, S., & Pratto, F. (1997). The interface between ethnic and national attachment. Ethnic
pluralism or etnic dominance? Public Opinion Quarterly, 61, 102–133.
22 Klandermans

Simon, B. (1999). A place in the world: Self and social categorization. In T. R. Tyler, R. M. Kramer, & O. P. John (Eds.), The
psychology of the social self (pp. 47–69). Mahwah, NJ: Lawrence Erlbaum.
Simon, B. (2011). Collective identity and political engagement. In A. Azzi, X. Chryssochoou, B. Klandermans, & B. Simon
(Eds.), Identity and participation in culturally diverse societies (pp. 137–157). Oxford, UK: Blackwell Wiley.
Simon, B., & Klandermans, B. (2001). Towards a social psychological analysis of politicized collective identity: Conceptu-
alization, antecedents, and consequences. American Psychologist, 56, 319–331.
Simon, B., Loewy, M., Stürmer, S., Weber, U., Kampmeier, C., Freytag, P., Habig, C., & Spahlinger, P. (1998). Collective
identity and social movement participation. Journal of Personality and Social Psychology, 74, 646–658.
Smith, H. J., & Tyler, T. R. (1996). Justice and power. European Journal of Social Psychology, 26, 171–200.
Staerklé, C., Sidanius, J., Green, E. G. T., & Molina, L. (2005). Ethnic minority-majority asymmetry and attitudes towards
immigrants across 11 nations. Psicologia Politica, 30, 7–26.
Stryker, S. (2000). Identity competition: Key to differential social movement participation? In S. Stryker, T. J. Owens, & R.
W. White (Eds.), Self, identity, and social movements (pp. 21–40). Minneapolis: University of Minnesota Press.
Stryker, S., Owens, T. J., & White, R. W. (2000). Self, identity, and social movements. Minneapolis: University of Minnesota
Press.
Stürmer, S., & Simon, B. (2009). Pathways to collective protest: Calculation, identification, or emotion? A critical analysis of
the role of group-based anger in social movement participation. Journal of Social Issues, 65(4), 681–705.
Tajfel, H., & Turner, J. C. (1986). The Social Identity Theory of intergroup behaviour. In S. Worchel & W. G. Austin (Eds.),
Social psychology of intergroup relations (pp. 7–24). Pacific Grove, CA: Brooks Cole.
Tarrow, S. (1998). Power in movement. Social movements, collective action and mass politics in the modern state. 2nd edition.
Cambridge, UK: Cambridge University Press.
Taylor, V., & Whittier, N. E. (1992). Collective identity in social movement communities: Lesbian feminist mobilization. In
A. Morris & C. Mueller (Eds.), Frontiers of social movement theory (pp. 104–130). New Haven, CT: Yale University
Press.
Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude-behavior relationship: A role for group identification.
Personality and Social Psychology Bulletin, 22, 776–793.
Turner, J. C. (1999). Some current issues in research on social identity and self-categorization theories. In N. Ellemers, R.
Spears, & B. Doosje (Eds.), Social identity (pp. 6–34). Oxford, UK: Blackwell.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group. A
self-categorization theory. Oxford: Blackwell.
Tyler, T. R., Boeckmann, R. R., Smith, H. J., & Huo, Y. J. (1997). Social justice in a diverse society. Boulder, CO: Westview
Press.
Van Dyke, N. (2003). Crossing movement boundaries: Factors that facilitate coalition protest by American college students,
1930–1990. Social Problems, 50, 226–250.
Van Stekelenburg, J. (2013). Collective Identity. In D. Snow, D. Della Porta, B. Klandermans, & D. McAdam (Eds.), The
Wiley-Blackwell encyclopedia of social and political movements (pp. 219–225). Malden, NJ: Wiley-Blackwell.
Van Stekelenburg, J., & Klandermans, B. (2007). Individuals in movements: A social psychology of contention. In B.
Klandermans & C. Roggeband (Eds.), Handbook of social movements across disciplines (pp. 157–204). Amsterdam,
Netherlands: Springer.
Walgrave, S., & Rucht, D. (2010). The world says no to war. Demonstrations against the war in Iraq. Minneapolis: University
of Minnesota Press.
Walsh, E. J. (1981). Resource mobilization and citizen protest in communities around Three Mile Island. Social Problems,
29(1), 1–21.
Zegeye, A., Liebenberg, I., & Houston, G. (N.d.). Resisting ethnicity from above: Social identities and the deepening of
democracy in South Africa. Pretoria: Human Science Research Council. Unpublished paper.
Copyright of Political Psychology is the property of Wiley-Blackwell and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.