“[…] yes, the fight against racism and discrimination must continue unabated. Fortunately, many different tools can be used to try and resolve these serious problems. However, none of them will work if one component is missing: laïcité. Not positive laïcité, not inclusive laïcité, not I-don’t-know-what laïcité, just laïcité, period. Only true laïcité, because it extols the universality of rights, makes it possible to exercise equality, liberty, fraternity and sorority. It alone permits true religious freedom, a freedom that every religion denies, more or less openly depending on how it is marketed, when faith stops being a purely personal affair and enters the political arena. Only it, ironically, allows both believers and non-believers to live in peace. Anyone who claims to defend Muslims by accepting the totalitarian religious discourse is, in fact, defending the oppressor. The first victims of Islamic fascism are Muslims” (editorial in Charlie Hebdo n 1178, published on January 14, 2015, the first issue after the terrorist attacks).

On January 7, 2015, terrorists opened fire in the offices of the satirical newspaper Charlie Hebdo, killing 11 people and wounding several others. By January 9, another two attacks had killed a female police officer at Montrouge and four people at a kosher supermarket. The strong emotions generated by these three attacks, all of which were carried out by people with links to Islamist groups, were felt not just by the French but by people around the world. During the days following the attacks, many “Je suis Charlie” demonstrations, as well as some few “anti Charlie” demonstrations, were held across the globe (see Mayer & Tiberj, 2016; Zerhouni, Rougier, & Muller, 2016). In France, the aftermath of the attacks saw a general clamor defending the country’s “Republican values”, most notably, freedom of expression and the principle of laïcité. This period also gave rise to numerous hate crimes against Muslims, with French law enforcement agencies recording 128 such acts between January 7 and 20, 2015, compared with 133 for the whole of 2014 (source – French Council of the Muslim Faith). A similar increase in hate crimes occurred in the United States following the September 11, 2001, terrorist attacks (Roehner, 2004). In fact, recent years have seen a worrying rise in prejudice and discrimination against Muslims in both Europe and the United States (Kaplan, 2006). For example, according to a 2015 report by France’s National Consultative Commission on Human Rights (CNCDH), in 2014, 45% of French people had a negative opinion of the Muslim faith. Nevertheless, despite this high percentage, fewer hate crimes were committed against Muslims in 2014 than in 2013. In contrast, the report noted a much lower percentage of negative opinions about the Jewish faith but a substantial increase in the number of anti-Semitic acts committed. A “flash survey” carried out in 2015 by CNCDH showed a slightly higher level of general tolerance of minorities in March 2015 compared with 2014. Based on this result, the CNCDH concluded that the terrorist attacks had not led to a durable “hardening of racist or anti-Semitic views” in France. In contrast, studies carried out following terrorist attacks in the United States in 2001, in Spain in 2004, and in the United Kingdom in 2005 suggest that these attacks led to greater prejudice against Muslims and against immigrants in general in these countries (Argyrides & Downey, 2004; Das, Bushman, Bezemer, Kerkhof, & Vermeulen, 2008; Echebbaria-Echabe & Fernadez-Guedes, 2006; Hitlan, Carrillo, Zarate, & Aikman, 2007; Morales-Marente, Moya, Palacios, & Willis, 2009; Van de Vyver, Houston, Abrams, & Vasiljevic, 2016). However, other studies have reported a number of positive social consequences of terrorist attacks, especially after the September 11 attacks (Scott Morgan, Wisneski, & Skitka, 2011), thereby highlighting the complexity of the effects produced by these extreme situations.

As Van de Vyver et al. (2016) noted, for obvious reasons the literature contains very few studies comparing data collected just before and immediately after a terrorist attack. The present paper reports data collected as part of a study carried out between December 3, 2014, and March 22, 2015. Our objective was to determine whether the January 2015 terrorist attacks were followed by an increase in prejudice toward immigrants and North-Africans, social dominance orientation (SDO) and attachment to the principle of laïcité.2

Social psychology research into the effects of terrorist attacks on the attitudes of the targeted countries’ citizens refers to several processes, the most important being activation of the idea of death and perceived threat. According to Pyszczynski, Solomon, and Greenberg (2002), because the September 11 terrorist attacks produced a large number of casualties and because they targeted one of the symbols of the United States, they activated and increased the salience of the idea of death. These authors suggest that one of the effects of this activation is to increase an individual’s attachment to the values of their ingroup. Similarly, Moskalenko, McCauley, and Rozin (2006) found that after September 11, 2001, individuals were more strongly attached to and identified more strongly with their ingroup. Finally, such events engender more negative attitudes towards people/groups seen as embracing opposing norms and values to those of the ingroup (Greenberg, Pyszczynski, & Solomon, 1986). In addition, the ability of terrorist acts to activate the idea of death favors an increase in prejudice against outgroups (Das et al., 2008; Kastenmüller, Greitemeyer, Ai, Winter, & Fischer, 2011), as does the perception of threat produced by a terrorist attack (Doosje et al., 2009; Hitlan et al., 2007). Terrorist attacks can also increase an citizens’ perceptions of a physical threat, especially to their own safety, and a symbolic threat to their identity, their values and their way of life (Doosje, Zimmermann, Küpper, Zick, & Meertens, 2009; Hitlan et al., 2007). The uncertainty-identity theory (Hogg, 2012; Hogg & Adelman, 2013) maintains that when individuals feel a threat to their safety or their way of life they are more likely to identify with a radical group they feel is capable of countering the threat and are globally less favorable toward policies of tolerance (Skitka, Bauman, & Mullen, 2004). They also tend to identify more strongly with their ingroup and have a greater preference for a clear hierarchy between social groups (Hogg & Adelman, 2013).

According to social dominance theory (Duarte, Dambrun, & Guimond, 2004; Pratto, Sidanius, Stallworth, & Malle, 1994; Sidanius & Pratto, 1999), individuals with a high SDO support the ideas of a clear social hierarchy between social groups, whereas individuals with low SDO prefer the idea of an egalitarian system between groups. SDO is an important determinant of intergroup prejudice (Guimond, Dambrun, Michinov, & Duarte, 2003; Pratto et al., 1994). Other studies have shown that SDO is affected by the perception of a threat to factors affecting the ingroup’s identity and that it can act as a mediating variable with respect to prejudice (Dambrun, Maisonneuve, Duarte, & Guimond, 2002; Guimond et al., 2003).

Taken together, this research suggests that terrorist attacks engender greater denigration of outgroups by citizens of the targeted country, who view these outgroups as a threat; a stronger preference for a clear social hierarchy; and an increase in the importance accorded to the ingroup’s defining norms and values. In light of these findings, we expected the January 2015 terrorist attacks to lead French people to show increased prejudice towards immigrants and North-Africans, increased SDO, and stronger attachment to the principle of laïcité. Indeed, the 1905 Act dictating the separation of church and state requires the French state to remain neutral with respect to religion, its role being to ensure freedom of religion and equal treatment of French citizens irrespective of their faith or absence of faith (Pena-Ruiz, 2014). Laïcité is a core value of the French Republic to which French people are strongly attached (Barthélemy & Michelat, 2007). Because the attack on the satirical newspaper Charlie Hebdo was an attack on both freedom of expression and the principle of laïcité, we expected the events of January 2015 to increase French people’s attachment to this principle.

An important issue when examining the effects of terrorist attacks on citizen’s perceptions and attitudes is the persistence of these effects. Research suggests that immediate reactions to terrorist attacks tend to be short term and that they fade when the events in question begin to lose their cognitive and emotional salience (Argyrides & Downey, 2004; Moskalenko et al., 2006, Roehner, 2004); however, not every study has confirmed this finding (e.g., Hitlan et al., 2007). The data used in the present research were collected just before the terrorist attacks and during the month following them. In order to examine the persistence of any effects produced by the attacks, we divided the data collected after the attacks into two groups covering two successive two-week periods. We examined three hypotheses:

H1: The January 2015 terrorist attacks would lead to an increase in prejudice toward immigrant and North-Africans, but this increase would be temporary and would only be observable during the two weeks immediately following the attacks.

H2: The increase in prejudice toward immigrant and North-Africans immediately after the terrorist attacks would be explained (i.e., mediated) by a temporary increase in SDO.

H3: The January 2015 terrorist attacks would lead to an increase in attachment to the principle of laïcité, but this increase would be temporary and would only be observable during the two weeks immediately following the attacks.

Method

Participants

We recruited participants from across France by sending an email to a network of acquaintances (who could forward the email to their networks) asking them to complete an online questionnaire. The resulting general population sample consisted of 162 people (Mage = 37.63; SD = 14.32).3 All participants were of French nationality, 39.50% were men and 59.88% were women (one person did not state their sex), 24.7% were students and 72.8% said they were in work. Employed participants reported a wide range of occupations (e.g., public employees, teaching, medicine, trade, construction, etc.). Seventy-four participants said they were believers (practicing or non-practicing, 45.7% of the sample), and 88 participants said they were non-believers (atheists and agnostics, 54.3% of the sample). We divided participants into three groups consisting of 65 participants (40.1% of the sample) who completed the questionnaire between December 3, 2014 and January 6, 2015 (first period), 56 participants (34.6% of the sample) who completed the questionnaire during the two weeks following the terrorist attacks (second period), and 41 participants (25.3% of the sample) who completed the questionnaire between two and four weeks after the terrorist attacks (third period). We divided participants in this way in order to place the January 7, 2015 terrorist attack in the middle of our data collection period. Hence, we obtained responses during the month preceding the attacks and during the month following the attacks. By dividing the post-attacks group into two consecutive two-week periods we were able to determine whether any of the attacks’ effects persisted over time. Table 1 summarizes the socio-demographic data for each group of participants.

Table 1

Socio-demographic data for the three data collection periods.

Period Gender Faith Socio-professional group

Period 1
n = 65
Women: 38 (58.5%)
Men: 27 (41.5%)
Non-believer: 37 (56.9%)
Catholic: 21 (32.3%)
Protestant: 2 (3.1%)
Muslim: 3 (4.6%)
Jewish: 1 (1.5%)
Pagan: 1 (1.5%)
Student: 18 (27.7%)
Employed: 50 (76.9%)
Period 2
n = 56
Women: 31 (55.4%)
Men: 25 (44.6%)
Non-believer: 14 (25%)
Catholic: 38 (67.9%)
Protestant: 2 (3.6%)
Buddhist: 1 (1.8%)
Christian: 1 (1.8%)
Student: 14 (25%)
Employed: 37 (66.1%)
Period 3
n = 41
Women: 28 (68.3%)
Men: 12 (29.3%)
Non-believer: 30 (73.2%)
Catholic: 9 (22%)
Deist: 1 (2.4%)
Not stated: 1 (2.4%)
Student: 8 (19.5%)
Employed: 31 (75.6%)

Measures

The questionnaire consisted in a number of scales presented in Likert format and ranging from 1 (“strongly disagree with the statement”) to 7 (“strongly agree with the statement”):

  1. Dambrun and Guimond’s (2001) scale of general prejudice towards immigrants and people of north-African origin consisting of 15 items (e.g., “I would not be worried if most of my friends at the university/at work were of Arab descent”, inverted item α = .91)
  2. Sidanius and Pratto’s (1999) SDO scale, translated into French by Duarte, Dambrun, and Guimond (2004), consisting of 10 items (e.g., “Some groups of people are simply inferior to other groups”, α = .82).
  3. A measure of attachment to the principle of laïcité adapted from Barthélemy and Michelat (2007) and consisting of 4 items (e.g., “Would you say you believe in the principle of laïcité”; “Could you tell me whether the word ‘laïcité has positive connotations for you”; “Do you feel that laïcité was worth fighting for but that today it is outmoded”, inverted item; “Do you feel that laïcité is still a relevant fundamental value”, α = .78). This measure allowed us to assess participants’ general opinions towards the principle of laïcité, without providing a definition of the principle.

Participants then had to say whether they were atheist, agnostic, a practicing believer or a non-practicing believer. Believers were asked to state which faith they believed in. The questionnaire ended with a series of demographic questions (e.g. age, sex, profession or year and subject of study, nationality).

Results

We began by testing the effect of the ‘sex’ and ‘religious belief’ variables on prejudice toward immigrants and North-Africans, SDO, and attachment to laïcité. A data collection problem prevented us from testing the ‘age variable’ and the large number of stated professions made it impossible to test the ‘socio-professional group’ variable. ANOVAs examining prejudice, SDO, and attachment to laïcité variables with respect to the participants’ sex (male vs. female) did not reveal any significant differences in prejudice, F(1, 161) = .43, p = .513, ηp2 = .003, or SDO, F(1, 161) = .23, p = .630, ηp2 = .001. However, we found a marginally significant effect on attachment to laïcité, F(1, 161) = 2.91, p = .090, ηp2 = .018, with attachment to laïcité being slightly greater among men (M = 5.16, SD = 1.75) than among women (M = 4.69, SD = 1.70). ANOVAs examining possible connections between religious belief (non believers vs. believers) and prejudice showed that the believers were significantly more prejudiced toward immigrants and North-Africans (M = 4.20, SD = 1.23) than the non believers (M = 3.25, SD = 1.26), F(1, 161) = 23.29, p < .001, ηp2 = .127. Furthermore, we found a marginally significant difference in SDO scores, with the believers having slightly higher scores (M = 2.47, SD = 1.23) than the non-believers (M = 2.14, SD = .98), F(1, 161) = 3.60, p = .060, ηp2 = .022. Our analysis of attachment to laïcité and religious belief did not reveal any differences between the two groups of participants, F(1, 161) = 1.70, p = .194, ηp2 = .011. These results led us to include sex and religious belief as covariables in all our subsequent statistical analyses.4 Table 2 presents means and standard deviations for each measure and for each data collection period.

Table 2

Means (and standard deviations) for prejudice toward immigrants and North-Africans, social dominance orientation (SDO) and attachment to laïcité for the three data collection periods.

Prejudice SDO Attachment to laïcité

Period 1 (n = 65) 3.49 (1.26) a 2.10 (1.03) 5.44 (1.56) a
Period 2 (n = 56) 4.32 (1.12) b 2.54 (1.33) 5.25 (1.80) a
Period 3 (n = 41) 3.12 (1.37) a 2.23 (.82) 3.44 (.98) b

Note: For each column, means with a different subscript differ at at least p < .05 according to post-hoc tests with Bonferroni correction.

Prejudice toward immigrants and North-Africans

An ANCOVA examining prejudice over the three data collection periods revealed a significant difference between the three periods, F(2, 160) = 6.69, p = .002, ηp2 = .08. A post-hoc test with Bonferroni correction showed levels of prejudice were significantly higher during the second data collection period than during the first (p = .022) and third periods (p = .002). Moreover, we did not find any difference in prejudice between the first and third data collection periods (p = .631). These results support our hypothesis that participants would show a temporary increase in prejudice toward immigrants and North-Africans immediately after the terrorist attacks.

Social dominance orientation

An ANCOVA examining SDO over the three data collection periods did not reveal any differences in SDO between the three periods, F(2, 160) = 1.60, p = .205, ηp2 = .02. This result is contrary to our hypothesis that the terrorist attacks would have an impact on the participants’ SDO.

Attachment to the principle of laïcité

An ANCOVA examining attachment to the principle of laïcité over the three data collection periods revealed a significant difference between the three periods, F(2, 160) = 26.59, p < .001, ηp2 = .25, due to a significant decrease in attachment to laïcité during the third period compared with the two preceding periods (comparison with periods 1 and 2, with Bonferroni correction: ps < .001). In addition, the post-hoc test did not show any difference between periods 1 and 2 (p = 1.00). Thus, contrary to our hypothesis, there was no significant increase in attachment to the principle of laïcité just after the terrorist attacks, but a decrease in this parameter between two and four weeks after the attacks.

This surprising finding led us to examine the correlations between our measures for each of the three periods. Results (see Table 3) revealed a substantial variation in the relations between the measures. In fact, although attachment to laïcité was negatively correlated with participants’ SDO during periods 1 and 2 (in line with results reported by Roebroeck & Guimond, 2014), attachment to laïcité was positively correlated with participants’ SDO during the third period. Hence, in the third data collection period, the higher the participants’ reported level of SDO, the higher their reported attachment to laïcité. These results suggest a change in the significance participants gave to their attachment to the principle of laïcité between the first two and the third measurement periods.

Table 3

Partial correlations for the three data collection periods between the measures of attachment to laïcité, prejudice toward immigrants and North-Africans and social dominance orientation (SDO), controlled for participants’ sex and religious belief.

Period 1 1 2 3

1. Attachment to laïcité
2. Prejudice –.07
3. SDO –.35 ** .34 **
Period 2 1 2 3

1. Attachment to laïcité
2. Prejudice –.30*
3. SDO –.54 *** .39 **
Period 3 1 2 3

1. Attachment to laïcité
2. Prejudice .14
3. SDO .36* .52 **

Note: *p < .05, **p < .01, ***p < .001. Correlations calculated without participants’ sex and religious belief as covariates gave similar results.

Discussion

Our objective for the present research was to examine the impact of the January 2015 terrorist attacks on prejudice against immigrants and North-Africans, SDO, and attachment to the principle of laïcité. We did this by comparing three groups of participants who completed a questionnaire either just before the terrorist attacks, during the two weeks following the attacks, or between two and four weeks after the attacks. A review of previous research (e.g., Doosje et al., 2009; Echebbaria-Echabe & Fernadez-Guedes, 2006; Van de Vyver et al., 2016) led us to expect an increase in prejudice against immigrants and people of north-African descent, SDO, and attachment to the principle of laïcité during the two weeks following the attacks (period 2) compared with before the attacks (period 1). However, participants were expected to get back to their initial levels of prejudice toward immigrants and North-Africans, SDO, and attachment to laïcité two to four weeks after the attacks (period 3).

Our data supported our hypothesis with respect to prejudice toward immigrants and North-Africans. In terms of the processes involved, the literature suggests that this type of increase in prejudice following a terrorist attack can be explained by the terrorist attack activating the idea of death (Das et al., 2008) and increasing perceived threat (Doosje et al., 2009). The links between these two processes require further investigation (Pyszczynski, 2015). The higher levels of prejudice toward immigrants and North-Africans recorded immediately after the attacks compared with the period between two and four weeks after the attacks suggest that the effects of the terrorist attacks were short lived; however, the exact cause of these effects remains uncertain. For example, experimental studies have shown that simply evoking past or fictitious terrorist attacks is enough to trigger an increase in prejudice (Das et al., 2008). The nature of the changes that occur in this type of situation also requires further study, because using self-reported measures of prejudice, as we did in the present study, allows participants to control, at least partially, the level of prejudice they express. Hence, rather than showing a real increase in prejudice, our results may simply reflect participants’ increased willingness to express long-held prejudices toward immigrants and North-Africans due to the terrorist attacks creating a normative context legitimizing the expression of these prejudices. This latter interpretation is in line with research showing differences in the expression of prejudice in different normative contexts (Crandall, Eshleman, & O’Brien, 2002). In the case of the January 2015 terrorist attacks in France, the resulting shock may have made it more socially acceptable to express certain prejudices, especially prejudice against people of north-African descent. This temporary change in the normative context could explain why people who are usually reticent about expressing certain prejudices may, in this context and for a short period, be more likely to express them. This type of change in normative context may also, at least in part, explain the temporary increase in hate crimes against people associated with a threat (e.g., Muslim) following terrorist attacks.

Our results did not show any significant variation in participants’ levels of SDO between the three data collection periods. SDO is an important determinant of an individual’s prejudice (Guimond et al., 2003; Pratto et al., 1994). Contrary to expectations, which were based on the uncertainty-identity theory (Hogg, 2012; Hogg & Adelman, 2013), there was nothing in our results to show that the terrorist attacks affected SDO or that the increase in prejudice toward immigrants and North-Africans was linked to a temporary increase in SDO. Another variable which may impact prejudice following terrorist attacks is right-wing authoritarianism (RWA; Altemeyer, 1981). According to the dual process model (Duckitt, Wagner, Du Plessis, & Birum, 2002), perceiving the world as competitive may increase intolerance towards outgroups by increasing SDO, while perceiving the world as dangerous may produce the same effect by increasing RWA. Hence, the January 2015 terrorist attacks may have had a greater effect on levels of RWA than on levels of SDO. Although our study did not allow us to test this supposition, it is supported by Echebbaria-Echabe and Fernandez-Guedes’s finding (2006) that the terrorist attacks in Madrid in 2004 led to an increase in authoritarianism as well as an increase in prejudice against outgroups. Further research is needed to examine this hypothesis more closely.

Also contrary to expectations, our analyses did not show an increase in participants’ attachment to the principle of laïcité during the period immediately after the terrorist attacks. Unexpectedly, we found a significant decrease in attachment to this principle during the third data collection period. The absence of an increase in attachment to the principle of laïcité just after the attacks is surprising given the political and media discourse during this period (Moliner, 2015). This result suggests that the principle of laïcité is not sufficiently clear and important to French people and that it is not linked to identification with the ingroup (Moskalenko et al., 2006). Although the decrease in attachment to the principle of laïcité during the third data collection period is difficult to interpret, the change in the relation between SDO and attachment to laïcité over the three data collection periods provides a number of clues to the processes involved. Indeed, laïcité is an ambiguous concept (Nugier, Oppin, Cohu, Kamieski, Robroeck, & Guimond, 2016; Nugier, Robroeck, Anier, Kleinlogel, Chatard, & Guimond, 2016) and its meaning may have changed across these different periods. Although SDO and attachment to laïcité were negatively correlated during the first two data collection periods, as they were in a study by Roebroeck and Guimond (2014), the correlation inverted during the third period. In other words, the stronger a participant’s feelings against the principle of equality and in favor of a hierarchy of social groups, the stronger their attachment to the principle of laïcité. This result suggests a possible change in the signification the participants gave to the notion of laïcité between the first two and the third data collection periods. Reporting the results of surveys showing strong ideological splits with respect to the “march for the Republic”, Moliner (2015) suggested that the terrorist attacks were followed by “conditions favoring the appearance of a false consensus regarding support for the march of January 11” (p.44). People with low SDO may have felt uncomfortable with this “false consensus” and rejected the principle of laïcité, which they felt was being instrumentalized after the attacks. Hence, like colorblind ideology (Knowles, Lowery, Chow, & Hogan, 2009), laïcité may function as a “malleable ideology”. The “march for the Republic” was followed by debate throughout French society about the notion of laïcité (Moliner, 2015). Some researchers believe that the laïcité introduced by the 1905 Act separating church and state is now being superseded by a “new laïcité” that restricts individual freedoms and that is particularly aimed at the Muslim faith (Baubérot, 2012; Hennette-Vauchez & Valentin, 2014; Roebroeck & Guimond, 2015). This “new laïcité” contrasts with France’s more “traditional” laïcité, which is based on the principles of equality and tolerance towards all religions (Baubérot, 2012). Hence, far from being unequivocal, the notion of laïcité may be associated with widely different beliefs and representations with potentially opposite political and social implications (Barthélemy & Michelat, 2007; Cohu, Maisonneuve, & Testé, 2016).

The combination of results showing a decrease in attachment to the principle of laïcité and an inversion in the relationship between attachment to the principle of laïcité and SDO during the third data collection period suggests that this decrease primarily concerned the most egalitarian individuals (i.e., those who expressed low SDO). Thus, it may be that these individuals distanced themselves from the notion of laïcité because of their discomfort with the falsely consensual discourse surrounding the “march for the Republic” (Moliner, 2015) and the increasingly powerful debate promoting this “new laïcité” (Alduy & Wahnich, 2015).

The conditions under which the present research was carried out imposed a number of methodological limitations on our study that reduce the scope of the results. For example, the quasi-experimental nature of our study meant we were unable to establish causal relationships between variables, and further research is needed to determine the social-psychological processes underlying our findings Nevertheless, our results represent a valuable contribution to the literature as very few studies have collected data from a general population sample before and after terrorist attacks. In addition to confirming the increase in prejudice immediately following a terrorist attack reported by previous studies (e.g., Echebbaria-Echabe & Fernadez-Guedes, 2006), our research indicates a number of avenues for further research. It highlights the fact that the effects of terrorist attacks can be complex, equivocal, and sensitive to contexts, measures, and time periods. Our results complement the CNCDH’s conclusions (2015) by showing that the January 2015 terrorist attacks had an immediate but short-lived effect on prejudice against immigrants and North-Africans. The attacks also impacted dimensions such as attachment to laïcité.