Prevalence and Impact of Hate Speech among Politicians in Switzerland

Abstract

Politicians may be particularly vulnerable to hate speech because of their public visibility and exposure to deviating opinions. They play a critical role in the functioning of a democratic system, and therefore, hate speech poses a potential threat because it can impede politicians’ freedom of expression or even lead politicians to resign from office. However, little is known about the prevalence and impact of hate speech targeting politicians. We therefore surveyed the politicians in the Canton of Zurich, Switzerland, 667 of whom participated (39.8% response rate). The prevalence of hate speech experiences was 29.7% in their total time as politicians and 20.6% in the past 12 months. Participants who had held a political office for longer and were affiliated with a political party on the poles of the political spectrum were more susceptible. Crucially, 29.4% of the affected politicians had contemplated resigning from politics due to hate speech. Thus, our study underlines a need for action. As the participants indicated they wanted to deal with hate speech as little as possible but still desired some form of intervention, political measures should strive for intervention options that require minimal contributions from the victims.

1. Introduction

Hate speech has manifold negative consequences [1]. For victims, it can lead to feelings of unsafety and fear, changes in lifestyle such as reduced social engagement, and perceived vulnerability [1,2,3]. For democratic societies that rely on free and public discourse, it can become a threat because it can impede politicians’ freedom of expression. To arrive at a prevailing solution to a social debate, it is essential to ensure the open expression of diverging opinions and perspectives. Hate speech might become particularly threatening to a democracy if politicians resign from office or if they refrain from openly expressing their perspectives because they have experienced hate speech. Critically, a politician’s relative prominence and association with certain thorny issues have been found to increase their risk of becoming a victim of hate speech [4].

In recent years, hate speech has received increasing academic interest both internationally and across disciplines [5]. Hate speech can be defined as the hostile expression of bias toward a group or one of its members based on an actual or presumed aspect of identity, such as gender, ethnicity, or social class [6,7,8]. Typically, hate speech is targeted at (presumed members of) groups that the perpetrator considers lower in social status [9] and is facilitated by anonymity, for example, on social media [10]. In the context of freedom of expression, hate speech can be viewed as both a descendant of and, more importantly, an opponent or obstacle to exercising this fundamental right [6].

Hate speech experiences are prevalent. A systematic review showed that up to 67% of young people in various countries were exposed to online and offline hate speech [11]. A representative survey of the general population in Switzerland indicated a prevalence of digital hate speech of almost 10% across an individual’s lifetime and approximately 5% in the past 12 months [12,13]. Prominent and public figures, such as politicians associated with controversial issues, might be particularly exposed to hate speech. However, not much is known about the prevalence of hate speech experiences and their consequences among politicians. A recent study in Germany examined social media hate speech targeting mayors, and over 50% of the respondents reported hostile actions [14]. Although a growing body of research underlines the issue of digital hate speech on social media (for recent reviews, see Castaño-Pulgarín and colleagues [15] or Matamoros-Fernández and Farkas [16]), hate speech goes beyond social media and should be investigated comprehensively.

However, some investigations have examined related, broader concepts of hate speech. In a survey conducted in Canada, approximately 30% of politicians reported having experienced some form of harassment [17]. In Sweden, online abuse, including direct threats, offensive comments, and comments associated with gender or sexuality, has been reported by more than 90% of members of parliament [18]. Cross-national surveys administered in the United Kingdom, Norway, New Zealand, and Queensland, Australia, found that approximately 80–90% of politicians had experienced some form of aggressive or intrusive behavior [19,20,21,22]. Similarly, Bjørgo and colleagues [23] investigated threats, hate speech, and troublesome incidents summarized as “unwanted incidents linked to political activity” among Norwegian politicians. The prevalence of unwanted incidents varied between 52% and 85% for different groups of politicians. When asked about the consequences of such unwanted incidents, approximately one in four reported examples of limited freedom of speech (e.g., hesitating to state a particular opinion). Crucially, between 16% and 28% of politicians on different political levels stated that they had considered giving up politics. Reports of such negative consequences have dramatically increased since the first survey was conducted in 2013 [23,24]. Similarly, Gorrell and colleagues [4] found that the number of abusive tweets that UK politicians received was associated with their decision to stand for re-election (or not).

Some of the aforementioned studies also investigated risk factors for hate speech susceptibility, with mixed results. They found no differences in susceptibility between female and male politicians [14,17,23]. However, other studies that investigated tweets about politicians reported a higher prevalence of hate speech experiences among women [25] and that female ethnic minority politicians were exposed to the highest level of threats [26]. Similarly, although neither Adams and colleagues [17] nor Bauschke and Jäckle [14] found significant age effects, Bjørgo and colleagues [23] reported reduced susceptibility with age. Lastly, party affiliation does not seem to affect susceptibility [14,17], but politicians at extreme ends of the political scale were found to be most affected [23].

Taken together, the existing empirical research has shown that politicians are vulnerable to negative experiences and that there is some evidence of a threat to free speech, indicating a potentially big impact of hate speech on politicians as well as democracy. The present study aimed to investigate all forms of hate speech among politicians, including those beyond social media, using survey data of politicians in Switzerland.

2. Materials and Methods

2.1. Participants and Procedure

All elected politicians in the Canton of Zurich, Switzerland, were invited to participate in the cross-sectional survey. They were either personally invited via email or their municipality. Participation was voluntary and anonymous; participants provided their informed consent by taking part in the survey. LimeSurvey was used as the online survey tool (LimeSurvey GmbH, Hamburg, Germany). Participants completed the survey online between 30 November and 19 December 2023.

2.2. Measures

The present study adopted a broad perspective of what constitutes hate speech to allow for a comprehensive assessment. The definition, which was also shown to participants at the beginning of the survey, was: Hate speech is a harmful expression targeting a personal characteristic, such as skin color, language, nationality, religion, gender, sexual identity, disability, physical appearance, gender expression, education, income, profession, or political view.

We created a questionnaire to investigate the prevalence and impact of hate speech among politicians (see

Table 1. Measures of hate speech experiences, their impact, and related concerns.

Measure	Question	Answer Options
All have been asked
Overall experience 1,2	Have you experienced hate speech since you became a politician?	“yes”, “no” 3
Twelve-months experience 1	Have you experienced hate speech in the past 12 months as a politician?	“yes”, “no” 3
Impact on being a politician	Do you feel affected in your role as a public official by hate speech?	“1-not at all affected” to “6-strongly affected” 3
Concerns	I refrain from filing a report for the protection of my family.	“1-fully disagree” to “6-strongly agree” 3
	I refrain from filing a report for my own protection.
	I would like the authorities to contact the authors.
	I would like the hate speech to be recorded, but without any further consequences.
	I want to deal with hate speech as little as possible.
	I want to spend as little time as possible for my safety and protection.
	I would like a criminal intervention.
	I would like a police intervention.
Demographic characteristics	Gender	“female”, “male”, “non-binary”, “no answer” 3
	Age	“under 20 years”, “20–29 years”, “30–39 years”, “40–49 years”, “50–59 years”, and “60 years or older” 3
	Political party affiliation	all popular parties in the Canton of Zurich as well as “independent” and “other” 3,4
	Time in current political position	“less than 1 year”, “1–2 years”, “3–5 years”, “6–10 years”, and “11 years or longer” 3
	Overall time as a politician	“less than 1 year”, “1–2 years”, “3–5 years”, “6–10 years”, and “11 years or longer” 3
Only those who had reported hate speech experience in the past 12 months have been asked
Impact on resigning	Did it ever occur to you to resign from your political position because of hate speech?	“yes, often”, “yes, sometimes” and “no” 3
Characteristic(s) targeted	Which personal characteristic(s) were targeted by hate speech?	“gender/gender identity”, “sexual orientation”, “religious affiliation”, “ethnicity/origin/skin color”, “disability”, “physical appearance”, “education/income/profession”, “political position/party”, and “other” 5
Channel	Through which channels have you experienced hate speech?	“personal”, “telephone”, “letter”, “email”, “SMS”, “messenger (e.g., WhatsApp)”, “Facebook”, “Instagram”, “Telegram”, “LinkedIn”, “TikTok”, “X (Twitter)”, and “other” 5
Number of experienced (a) incidents and (a) reactions within the past 12 months 6

¹ Used to calculate the prevalence of hate speech. ² Used as an outcome in logistic regression. ³ Answer format: single answer. ⁴ Party affiliation was categorized as “left-leaning”, “centrist”, “right-leaning”, and “independent” based on the official classification of the Canton of Zurich [27]. ⁵ Answer format: multiple answers; ⁶ used to calculate the ratio of hate speech experiences to which the participants reacted.

). Specifically, we assessed the experience of hate speech, both overall and within the past twelve months, its impact on being a politician, concerns related to hate speech, and demographic characteristics. Those who had reported a hate speech experience in the past 12 months were further asked about targeted characteristic(s), channel(s) through which hate speech was received, number of incidents and reactions, and impact on resigning. The questions were presented in German, but English translations are reported in Table 1.

2.3. Statistical Analysis

The pre-processing of data and all subsequent statistical analyses were carried out using R Studio version 2021.09.0 and R version 4.1.1 [28]. Pairwise deletion was chosen to handle missing data, resulting in different numbers of participants for different items/analyses. However, percentages are always expressed as a ratio of the whole population of participants (i.e., not just of those who answered each question; N = 660).

Bivariate logistic regression models were fitted to investigate associations between overall hate speech experience and risk factors. Risk factors were gender, age, political party affiliation, and length of time as a politician. Corresponding reference categories were female, highest age group, no political party affiliation (independent), and longest timespan as a politician. Gender was dichotomized as female vs. male for this analysis because only one participant selected “non-binary”. Similarly, the two youngest age groups were combined into “30 years or younger”, and the two lowest categories of time as a politician were combined into “2 years or less”.

3. Results

3.1. Sample Characterization

All elected politicians (N = 1676) in the Canton of Zurich, Switzerland, were invited to participate in the survey. Of the 1456 personally invited ones, 634 (43.5%) decided to participate in the survey. Of the 220 politicians invited via their municipality, 33 (15%) participated in the survey. Thus, 667 politicians participated in the study (39.8% participation rate). Seven participants were excluded from the analyses because they agreed to participate but did not provide any answers. Therefore, the analysis sample includes 660 politicians (39.4%). A missing data analysis using a chi-squared test showed that participation was not significantly associated with gender (derived from the political position, which is gendered in German). Most of the 660 politicians were male, 50 years or older, and had spent 6 or more years in office as well as in their current political position. Additional details can be found in

Table 2. Characteristics of participating politicians (N = 660).

Variable	% (n)
Gender
Female	31.6 (209)
Male	57.9 (382)
Non-binary	0.2 (1)
No answer	10.3 (68)
Age group
<20 years	0.2 (1)
20–39 years	3.2 (21)
30–39 years	8.0 (53)
40–49 years	20.3 (134)
50–59 years	33.6 (222)
≥60 years	25.0 (165)
No answer	9.7 (64)
Political party
Left-leaning	17.7 (117)
Centrist	15.2 (100)
Right-leaning	30.8 (203)
Independent	24.5 (162)
No answer	11.8 (78)
Length of time as a politician
<1 year	2.6 (17)
1–2 years	14.7 (97)
3–5 years	12.6 (83)
6–10 years	22.1 (146)
≥11 years	36.4 (240)
No answer	11.7 (84)
Length of time in current position
<1 year	4.4 (29)
1–2 years	29.7 (196)
3–5 years	18.2 (120)
6–10 years	25.2 (166)
≥11 years	12.7 (84)
No answer	9.8 (65)

.

3.2. Prevalence of Hate Speech and Risk Factors

The overall prevalence of experiencing hate speech as a politician was 29.7% (0.5% no answer). The 12-month prevalence was 20.6% (1.0% no answer). The results of the bivariate logistic regression analyses are presented in

Table 3. Results of bivariate logistic regression models investigating the risk factors for overall hate speech experiences.

Independent Variable	N	Odds Ratio [2.5%, 97.5%] *	p-Value
Gender (reference: female)	591
Male		0.78 [0.54, 1.23]	0.19
Age group (reference: ≥60 years)	596
<30 years		1.67 [0.63, 4.20]	0.28
30–39 years		1.64 [0.84, 3.16]	0.15
40–49 years		1.16 [0.69, 1.94]	0.57
50–59 years		1.11 [0.70, 1.76]	0.66
Political party (reference: independent)	582
Left-leaning		3.90 [2.30, 6.75]	<0.001
Centrist		1.17 [0.62, 2.17]	0.62
Right-leaning		1.76 [1.07, 2.93]	0.03
Length of time as a politician (reference: ≥11 years)	583
≤2 years		0.41 [0.23, 0.72]	0.002
3–5 years		0.65 [0.36, 1.15]	0.15
6–10 years		1.25 [0.81, 1.93]	0.31

* 95% confidence interval of the odds ratio.

. Neither gender nor age group showed a significant effect. However, party affiliation and length of time as a politician were significantly associated with overall hate speech experience: politicians affiliated with a left- or right-leaning political party had significantly higher odds of having experienced hate speech compared with independent politicians. Additionally, participants who had been politicians for 11 years or longer were significantly more likely to have experienced hate speech compared with politicians with two years of experience or less.

3.3. Impact of Hate Speech Experiences

Out of all the participants, 87.7% stated that they did rather not, not, or not at all feel affected by hate speech as public officials. Conversely, 5.9% reported feeling at least somewhat affected by hate speech, and 6.2% did not provide any response. Of the participants who reported experiencing hate speech in the past 12 months, 14.7% felt rather affected (15.4% did not answer). Also, 29.4% of the participants who experienced hate speech in the past 12 months reported that they had thought about resigning from their political position due to the hate speech they received. Of these participants, 26.5% answered that they considered it “sometimes”, 2.9% considered it “often”, and 7.4% did not specify further.

3.4. Characterization of Hate Speech Experiences

Only participants who reported experiencing hate speech in the past 12 months were asked to provide further information about their experience(s). Detailed results can be found in

Table 4. Characterization of hate speech experiences within the last 12 months.

Variable		% (n)
Characteristic targeted by hate speech
	Political position/party	86.0 (117)
	Education/income/profession	22.1 (30)
	Gender/gender identity	18.4 (25)
	Physical appearance	11.0 (15)
	Sexual orientation	6.6 (9)
	Religious affiliations	3.7 (5)
	Ethnicity/origin/skin color	3.7 (5)
	Disability	0 (0)
	Other	11.0 (15)
	No answer	2.2 (3)
Channel through with hate speech was received
	Email	46.3 (63)
	Personal	44.1 (60)
	Facebook	27.9 (38)
	X (formerly Twitter)	24.3 (33)
	Letter	21.3 (29)
	Telephone	9.6 (13)
	Instagram	8.1 (11)
	Messenger (e.g., WhatsApp)	5.9 (8)
	SMS	2.2 (3)
	Telegram	1.5 (2)
	LinkedIn	0.7 (1)
	TikTok	0.7 (1)
	Other	11.0 (15)
	No answer	4.4 (6)

. The most frequently reported characteristic targeted by hate speech was their political position or party, followed by some distance by their education, income, or profession and their gender or gender identity. Email and personal contact were the most frequently reported channels through which hate speech was received. However, the combined percentage for all social media channels 1 was 46.3%. The median number of hate speech experiences within the past 12 months was 3 (interquartile range [IQR]: 2–10, 3 missing). Most participants reported that they did not react to any of their experiences (median: 0, IQR: 0–33.33, 3 missing).

3.5. Potential Concerns Regarding the Handling of Hate Speech

Figure 1 presents the ratings of the respondents’ potential concerns about the handling of hate speech (full sample). The highest proportion of agreement was found for the concerns “I want to spend as little time as possible for my safety and protection” and “I want to deal with hate speech as little as possible”, followed by “I would like the authorities to contact the authors” and the wish for a police or criminal intervention.

4. Discussion

Politicians are responsible for expressing and discussing political arguments in the public eye. Therefore, they might be especially susceptible to hate speech, which has manifold negative consequences directly on them and society. Hate speech can lead them to hesitate to express certain opinions or viewpoints, thus limiting freedom of speech, or even to resign from office. Therefore, hate speech poses a threat to politicians and the functioning of democracies. In this study, new insights about the prevalence and impact of hate speech among politicians were gained by surveying 667 politicians in Switzerland, a country with the most direct form of democracy. We obtained three main results.

First, hate speech was found to be prevalent among Swiss politicians: one in three participating politicians reported having experienced hate speech in all their time as a politician, and one in five reported the same for the past 12 months. These prevalence estimates are considerably higher than those reported by Stahel and colleagues [12,13] for the general population in Switzerland, which were approximately one in 10 (lifetime) and one in 20 (in the past 12 months). Although the rates are not directly comparable because Stahel and colleagues [12,13] restricted their investigation to digital hate speech, the hypothesis that politicians are especially susceptible to hate speech seems empirically plausible. For more specific conclusions, a comparative study including politicians as well as other potential target groups of hate speech would be required. Additionally, our results show that a significant proportion of hate speech is (still) transmitted personally or via letters, thereby underlining the neglect of studies that only investigated digital forms. Nevertheless, the prevalence of hate speech among Swiss politicians appears to be rather low by international comparison. The only other study that investigated hate speech targeting politicians was conducted in Germany and found a prevalence of approximately one in two [14]. Again, the rates are not directly comparable due to the different scopes of the surveys. Still, because the German study investigated only digital hate speech, one would expect the rate in our survey (of all forms of hate speech) to be higher if the prevalence of all hate speech experiences were, in fact, comparable in the two surveys. Similarly, though even more difficult to compare due to their investigation of concepts broader than hate speech, surveys in Norway, the UK, New Zealand, and Queensland, Australia showed prevalence estimates between 50% and 90% [21,23,24]. Only one study from Canada reported results similar to ours, with one in three politicians experiencing some form of harassment [17].

Second, our research revealed that politicians who are associated with right- or left-leaning parties are more prone to experiencing hate speech compared with politicians registered as “independent”. In contrast, politicians belonging to centrist parties showed a similar prevalence of hate speech experiences as those with no party affiliation. Therefore, party affiliation at the margins of the political spectrum was a risk factor for hate speech experiences. This is consistent with Bjørgo and colleagues [23,24], who reported that politicians at the extreme ends of the political spectrum were more exposed to hate speech than those closer to the center. Gender and age were not determining factors in susceptibility to hate speech experiences, but politicians who had been in office for a longer period were more likely to have experienced hate speech. This might be because more time allows for more opportunities for hate speech incidents. Considering the political party and time spent in office, the odds of a hate speech incident were higher for politicians affiliated with a right- or left-oriented party for a longer time.

Third, almost one in three people who experienced hate speech reported having considered resigning from politics. This result is somewhat surprising because the rate is higher than that reported by Bjørgo and colleagues [23], even though their reported prevalence of unwanted incidents (including hate speech) was two to almost three times higher than the prevalence of hate speech found in the present survey. Hate speech incidents in Switzerland may be perceived as more impactful due to their relative infrequency, thus attributing greater significance to each occurrence. Future (international) studies should follow up on this finding. Additionally, a more qualitative approach to investigating former politicians might shed more light on our finding of the high rate of considering resigning from politics. Regardless of this finding, most participants reported that they never reacted to their hate speech experiences. In accordance, the concerns with the highest agreement rate were about dealing as little as possible with hate speech, although the next highest agreement rates were found for concerns related to some form of intervention.

4.1. Limitations

The participation rate of 40% was comparable to those of previous surveys of politicians, which showed rates between 34% and 60% [14,17,21,23]. However, the generalizability of the results to all politicians in the Canton of Zurich or even other regions might be limited due to the unknown representativity of the sample. Although gender was not significantly associated with participation in our survey, indicating that there was no selection effect regarding gender, we cannot exclude that other variables may have played a role. Our sample involved relatively few young politicians, rendering potential age effects difficult to detect. Future studies should further investigate how sociodemographic characteristics including minority status are associated with susceptibility to hate speech experiences in politicians as compared with the general population since risk factors do not seem to align between those groups. Additionally, investigating a more fine-grained left–right scale might lead to more insights into the effects of political party association.

4.2. Conclusions

Politicians are susceptible to hate speech experiences. This might affect the functioning of democracies as one in three affected politicians had contemplated resigning from politics. Our results call for political measures addressing the hate speech experiences of politicians. Importantly, interventions should demand as little involvement as possible of the victims, as they do not want to engage themselves.