Predictors of Street Harassment Attitudes in British and Italian Men: Empathy and Social Dominance

Abstract

Background: Street harassment is a widespread issue affecting women’s freedom and well-being. While research has explored its impact on victims, little is known about individual factors influencing societal tolerance of such behaviour. This study examined the relationships of cognitive empathy and social dominance orientation (SDO) to street harassment tolerance among heterosexual men in two European cultures. Methods: A total of 136 Italian and 113 British heterosexual men completed measures of SDO and street harassment tolerance before and after watching a video depicting a woman experiencing harassment. Cognitive empathy was assessed through participants’ ability to recognise the victim’s emotional responses in the video. Results: Higher tolerances of street harassment before and after viewing the video were associated with lower cognitive empathy and higher SDO in both countries. Exposure to the video reduced street harassment tolerance in British participants but had no significant effect in the Italian sample. Conclusions: These findings highlight cognitive empathy and SDO as correlates of street harassment tolerance, suggesting their potential relevance for interventions to shift street harassment attitudes. The differing impact of the video between countries highlight the potential importance of cultural context for developing street harassment prevention strategies. Future research should further explore these relationships to inform targeted interventions.

1. Introduction

1.1. Street Harassment: Definition, Prevalence, and Consequences

Street harassment involves verbal, non-verbal, or physical unwanted behaviour perpetrated by strangers in public spaces. Various forms of street harassment occur in public (e.g., racist, homophobic, and other discriminatory behaviours), all of which are social problems in need of further research. This study focuses on sexual harassment perpetrated by heterosexual male strangers against women, which typically involves unsolicited expressions of sexual appreciation or advances.

The motivations, manifestations, and impacts of this type of street harassment can differ from other types, warranting focused examination (Fileborn & O’Neill, 2023; Fischer & Good, 1994). While “street harassment” can take many forms, in this paper, we use the term specifically to refer to sexual harassment perpetrated by male strangers against women.

The high frequency of street harassment compounds its social impact. It is arguably the most common form of sexual harassment (Macmillan et al., 2000). Gekoski et al. (2017), for example, report widespread harassment on public transport, with prevalence rates ranging from 15% in the UK to 98% in Egypt. A recent international survey of over 15,500 women from 15 countries revealed that 80% had experienced street harassment at least once (International Survey on Sexual Harassment in Public Spaces, 2021).

Exposure to street harassment negatively impacts women’s lives in several ways. It diminishes well-being, increasing feelings of fear, anger, and violation (Bastomski & Smith, 2017; Betts et al., 2019; Kearl, 2014; Lenton et al., 1999; Macmillan et al., 2000; Nielsen, 2002). Victims report higher rates of depression, anxiety, and poor sleep quality (Betts et al., 2019; DelGreco & Christensen, 2020), and they are more likely to self-objectify themselves (Fairchild & Rudman, 2008; Fredrickson & Roberts, 1997). Street harassment heightens women’s fear of sexual violence, leading to constrained freedom of movement and behavioural changes, such as constant safety assessments and avoidance of certain places (Bastomski & Smith, 2017; Hickman & Muehlenhard, 1997; Lenton et al., 1999; Vera-Gray & Kelly, 2020). Additionally, women who experience or witness street harassment exhibit stronger negative intergroup emotions towards men (Chaudoir & Quinn, 2010) and a heightened desire for justice against perpetrators (Fileborn & Vera-Gray, 2017).

1.2. Street Harassment: Literature and Limitations

Although recent research has illuminated the impact of street harassment on victims, the factors contributing to the tolerance and endorsement of such behaviour remain under-researched. Street harassment tolerance refers to the degree to which individuals or societies perceive street harassment as acceptable (Vogt et al., 2007). This construct is closely related to sexual harassment tolerance, where attitudes towards harassment in broader contexts (e.g., workplaces) have been linked to the prevalence of such behaviours (Russell & Trigg, 2004). Understanding street harassment tolerance is crucial as reducing harassment requires addressing not only perpetrators but also societal norms that tacitly permit such behaviour.

Perpetrators of street harassment often remain unidentified due to the brief nature of encounters and anonymity. This makes the direct study of motivations and cognition of perpetrators particularly challenging. Studying predictors of societal tolerance of street harassment provides an indirect way to understand its psychological roots. Men who are more tolerant of street harassment and who are surrounded by peers sharing similar views are more likely to engage in such behaviour (Darnell & Cook, 2009). More broadly, attitudes towards rape, sexual harassment, and street harassment predict engagement in these behaviours (DelGreco et al., 2021; Lanier, 2001; Reilly et al., 1992).

A growing body of literature has explored social and attitudinal predictors of harassment tolerance. For example, Moya-Garófano et al. (2022) examined how ambivalent sexism influences women’s reactions to street harassment, while Saunders et al. (2017) found that hostile sexism among men affects how they expect women to cope with stranger harassment. Similarly, Russell and Trigg (2004) identified hostile sexism as a key factor in both men’s and women’s tolerance of sexual harassment. Despite these findings, cognitive and attitudinal predictors of street harassment tolerance specifically remain underexplored.

The effectiveness of video-based interventions on street harassment has been mixed (Darnell & Cook, 2009). However, recent social media campaigns featuring footage of street harassment have effectively increased awareness and promoted discussion, which can be tested as a potentially empathy-evoking intervention for reducing street harassment tolerance.

1.3. Cognitive Empathy, Social Dominance Orientation, and Population Differences

Empathy, the tendency to understand and share others’ feelings, has cognitive and affective components (Blake & Gannon, 2008). Cognitive empathy refers to the ability to understand others’ emotional states, while affective empathy involves sharing those emotions. This study focused on cognitive empathy, which is more directly linked to understanding the victim’s perspective—which could be a key factor in predicting harassment tolerance. While affective empathy was implicated in Darnell and Cook’s (2009) study, our focus on cognitive empathy stemmed from its established association with aggressive behaviours (Miller & Eisenberg, 1988) and the belief that understanding a victim’s distress could deter harassing behaviour.

A meta-analysis by Jolliffe and Farrington (2004) found that general criminal offending is associated with lower cognitive empathy. Similarly, lower cognitive empathy correlates with higher acceptance of myths that justify harassment (Diehl et al., 2014). DelGreco et al. (2021) found that belief in the victim’s negative experience reflects higher cognitive empathy. Therefore, the first aim of this study was to test whether lower cognitive empathy correlated with higher street harassment tolerance.

Another potential contributor is social dominance orientation (SDO), which reflects an individual’s support for social hierarchies and dominance over lower-status groups (Pratto, 1994). SDO has been linked to various forms of discriminatory behaviour, including sexual harassment (Pryor, 1994). Those with higher SDO may engage in street harassment to assert dominance, objectify others, or maintain social power (Berdahl, 2007; Halper & Rios, 2019). DelGreco et al. (2021) found that men feeling less powerful than women were more likely to harass, suggesting that SDO could underlie this dynamic. Thus, the second aim of this research was to explore whether higher SDO was associated with increased street harassment tolerance.

Cultural influences on street harassment attitudes also remain underexplored, yet they are vital for developing context-specific interventions. While our study sought to replicate findings across two Western European countries, Italy and the United Kingdom (UK), we also recognised the potential influence of distinct gender norms and cultural attitudes. Research indicates that Southern European countries like Italy retain more traditional gender values compared with Northern European countries like the UK (Bosoni & Baker, 2015). For example, a Gallup survey (Crabtree & Nsbuga, 2012) revealed that 40% of women in Italy feel safe walking alone at night, compared with 62% in the UK. Such differences suggest that street harassment tolerance may vary across cultures, which may be due to differences in culture-level cognitive empathy and SDO. By examining these factors in two culturally distinct populations, we aimed to explore both universal patterns and culturally specific dynamics.

1.4. Current Study

In summary, this study aimed to address significant gaps in street harassment research by examining the relationship of cognitive empathy and SDO to street harassment tolerance. We tested the following hypotheses:

(a)

Lower cognitive empathy for harassed women predicts higher street harassment tolerance.

(b)

Higher SDO predicts higher street harassment tolerance.

Additionally, we conducted the study in two culturally distinct European countries, in Italy and in the UK, to assess the replicability of findings and the relationship of cultural context to street harassment tolerance.

Finally, we examined whether tolerance of street harassment would change pre- and post-exposure to a video depicting street harassment and whether these changes in tolerance were associated with shifts in cognitive empathy and SDO scores.

We used multiple regression analyses to examine the unique contributions of cognitive empathy and SDO, allowing us to control for potential confounding factors and explore possible interactions between these constructs.

Through this research, we aimed to contribute to the understanding of cognitive and social factors related to street harassment tolerance and to inform interventions targeting these attitudes across diverse cultural contexts.

2. Materials and Methods

We used a mixed-methods design, combining correlational analysis to examine associations between cognitive empathy, SDO, and street harassment tolerance, with an experimental pre-/post-analysis to assess changes in street harassment tolerance following video exposure.

2.1. Participants

The inclusion criteria required participants to be male and over 18 years old. Seven responders identified as non-binary or another gender or preferred not to say and were excluded. As the video showed sexual comments perpetrated towards an individual of the opposite sex, only heterosexual individuals were retained in the study, excluding an additional 23 responders. The final two samples in this study were Italian (n = 136) and British (n = 113) individuals who self-identified as men and heterosexual. The participants’ ages ranged from 18 to 77 years in the Italian sample (M = 34.2, SD = 12.5) and 19 to 68 years in the British sample (M = 28.7, SD = 9.3). Participants were recruited through Prolific and social media. Prolific recruitment targeted 80 Italian and 80 British male participants, who were compensated GBP 2.70 for their time. Additional participants were recruited via social media posts in Italian and British community groups on Facebook. These participants were not compensated. Information about the participants’ highest education achieved is depicted in

Table 1. Descriptive table for education grouped by nationality.

Highest Education Level	Italians	British	Total
High School	50	77	127
Degree	28	45	73
Master	22	17	39
Ph.D.	9	1	10

. Information about their relationships can be found in Appendix A,

Table A1. Relationship information of samples.

	Italian	British
Ever had a romantic relationship with a female
Yes	102	93
No	28	18
Rather not say	6	2
Ever had a meaningful relationship with a female
Yes	125	105
No	9	7
Rather not say	2	0
Gender composition of friendship group
Entirely male	8	7
Mostly male	52	41
A mix of gender	59	61
Mostly female	17	4
Entirely female	0	0

.

2.2. Online Questionnaire

The questionnaire was delivered via a Qualtrics (Provo, UT, USA) online survey and was available in English and Italian to ensure comprehension. The English version was translated into Italian by a bilingual individual and back-translated by another bilingual to ensure consistency. Minor discrepancies were resolved collaboratively.

SDO-7 Scale. We used the SDO-7 scale (Aiello et al., 2019; Ho et al., 2015), an adaptation of Pratto (1994) original version. The scale contained 16 items rated on a 7-point Likert scale (1 = “strongly oppose”, 7 = “strongly favour”), with eight items reflecting social dominance (e.g., “An ideal society requires some groups to be on top and others to be on the bottom”) and eight reflecting anti-egalitarianism (e.g., “It is unjust to try to make groups equal”). All items were combined into a single score, following Ho et al.’s (2015) guidance for calculating an overall SDO score. This decision reflected the high internal consistency of the unified scale (Cronbach’s alpha = 0.91 across both samples). Eight items were reverse-coded to mitigate response biases. Higher scores indicated higher levels of SDO.

Street Harassment Tolerance Scale. This 11-item scale was modelled after Darnell and Cook’s (2009) measure, based on Goodchilds and Zellman’s (1984) sexual aggression acceptance scale. The items were rated on a 5-point Likert scale (1 = “not at all acceptable”, 5 = “very acceptable”). In Darnell and Cook’s (2009) scale, all items began with a common incipit: “How acceptable do you think it is for a man to make an unsolicited, unreciprocated, and unnecessary comment toward an unknown woman on the street (for example, saying ‘hey baby’ or ‘nice ass’) when”. Then, each item concluded with one of the following situations: the woman is attractive, the woman is dressed in sexy clothing (e.g., short skirt, tight clothes), the woman makes eye contact with him, the woman smiles at him, the woman is alone, the woman is with her friends, the woman is with a man, the woman is with her children, the man is alone, the man is with his friends, and the man is in an unfamiliar neighbourhood. We slightly modified the original items by removing “unreciprocated and unnecessary” to avoid leading participants towards a particular interpretation of the behaviour. The terms “unreciprocated” and “unnecessary” carry an implicit negative connotation, which could suggest to participants that these behaviours are inherently wrong or undesirable. By removing these terms, we aimed to reduce any bias in participants’ responses, allowing them to consider the acceptability of street harassment without being influenced by pre-defined negative judgments. Higher scores indicated greater tolerance. Cronbach’s alpha was 0.96 across both samples.

Video Stimulus. The video, sourced from Rob Bliss’s YouTube account (Bliss, 2014), depicts real footage of a woman walking through New York City while experiencing verbal and non-verbal street harassment. The video lasts two minutes, and the protagonist does not respond to or encourage the harassers. The final part of the video, which includes a message quantifying the instances of harassment and references to support services, was not shown to participants to avoid eliciting socially desirable responses. The video can be accessed at https://www.youtube.com/watch?v=b1XGPvbWn0A (accessed 5 October 2021).

Cognitive Empathy Scale. This 10-item self-report questionnaire was adapted from Fernandez and Marshall’s (2003) Rapist Empathy Scale and was previously used by Darnell and Cook (2009) to measure cognitive empathy in the context of street harassment. Darnell and Cook (2009) demonstrated that this scale effectively captured women’s experiences of street harassment, drawing from both past research (Lenton et al., 1999) and anecdotal data (Darnell & Cook, 2009). Unlike the original version, which referenced a vignette, our adaptation framed the items in relation to the video shown during the study. Each item followed the common stem, “How much do you think the woman in the video was feeling…”, concluding with one of ten emotions—five positive (complimented, proud, self-confident, pleased, safe) and five negative (offended, ashamed, guilty, afraid, angry). The items were rated on a 10-point Likert scale (1 = “not at all”, 10 = “very much”). Scores for positive and negative emotions could be computed separately to form two sub-variables: cognitive empathy for positive emotions and cognitive empathy for negative emotions. Previous research has combined positive and negative items to create a single composite measure of cognitive empathy (Darnell & Cook, 2009; Fernandez & Marshall, 2003) and the high internal consistency observed across both subscales in the current study (Cronbach’s alpha = 0.81), consistent with previous findings. Our analysis assumed that reality was both a low positive and highly negative experience for the harassment victim in the video, so we similarly combined them into a single composite cognitive empathy score (with lower positive and higher negative scores (reverse coded for composite) indicating greater empathy).

2.3. Procedure

The study was approved by the School of Psychology Ethics Board (application ID: ER/AG711/2) and complied with British Psychological Society ethical guidelines. Participants completed the survey remotely via computer or mobile devices. Informed consent was obtained, and participants were advised of their right to withdraw at any point. They were informed that the study focused on social interactions between men and women and warned about the potentially distressing content of the video. Participants were told they could exit the survey at any time if they felt uncomfortable. After completing a CAPTCHA, participants provided demographic information, including age, gender, sexual orientation, education level, relationship status, friendship group, and gender composition. The latter three were used for exploratory analyses for undergraduate dissertations and to describe our sample but were not a focus of the present study. Participants then completed the SDO and Street Harassment Tolerance scales in random order. Following this, they viewed the video stimulus. After the video, participants completed the Cognitive Empathy Scale and repeated the Street Harassment Tolerance Scale to assess post-video attitudes. All scales were presented in blocks, and participants could only proceed after completing all items and watching the entire video. No time limits were imposed to allow participants to complete the survey at their own pace. At the end of the study, participants were debriefed, with a detailed explanation of the study’s purpose and goals. They were reminded of their right to withdraw their data, and they were provided with contact information for mental health support services should they wish to access them. The debrief also included a reminder that participants could contact the research team if they had any questions or concerns following their participation.

2.4. Analysis Plan and Statistical Power

To analyse the data, we employed Tobit regression models to account for the floor effects observed in street harassment tolerance scores as many participants reported minimal tolerance, clustering at the lower end of the scale. Tobit regression models are commonly used when the dependent variable is censored, meaning there is a limit to the values it can take (in this case, a lower bound for tolerance scores). This model allows for the inclusion of these censored observations while providing unbiased parameter estimates under such conditions. Associations between cognitive empathy (positive, negative, and composite scores), SDO, and street harassment tolerance were examined. Pre- and post-video tolerance scores were analysed separately to evaluate the immediate impact of the video stimulus. Model selection was based on the Akaike information criterion (AIC) to identify the best-fitting models. A lower AIC value indicated a better model fit while penalising model complexity.

Given the number of analyses, we conducted an a priori power analysis using G*Power. With an alpha level of 0.05, a power of 0.80, and a medium effect size (f² = 0.15), a sample size of 107 participants per group was sufficient to detect significant effects. Our final samples exceeded this threshold, ensuring sufficient power for detecting medium-sized effects. Missing data were minimal (<2%) and were handled using listwise deletion as the pattern of missingness was random.

3. Results

3.1. Impact of Video on Street Harassment Tolerance

All participants were included in the analysis. Paired-sample t-tests were conducted to assess changes in street harassment tolerance before and after the video. The video significantly reduced street harassment tolerance in the British sample (t(111) = −3.93, p < 0.001, d = −0.37), indicating a small-to-medium effect size. No significant change was observed in the Italian sample (t(131) = −1.00, p = 0.319, d = −0.087), suggesting that the video had little to no impact. Pre- and post-video tolerance scores were highly correlated in both samples (r > 0.70), indicating stability in attitudes over time.

Table 2. Descriptive statistics for the study’s variables grouped by nationality.

		Italian		British
	Range	M	SD	M	SD
Street harassment tolerance	11–55	16.76	8.01	18.88	7.71
Cognitive empathy	10–100	77.05	13.21	68.91	11.64
Social dominance orientation	16–112	61.64	7.96	68.30	7.96

shows the descriptive statistics for participants’ street harassment tolerance, SDO, and cognitive empathy after watching the video.

3.2. Psychological Predictors of Street Harassment Tolerance

Street harassment tolerance responses had a floor effect, censored at a score of 11. It was predicted that if scores could go more negative than “not at all acceptable” in terms of opinions of street harassment, we would have a more Gaussian distribution of responses. We constructed Tobit regression models (VGAM package in R, v 1.1-8) to consider this lower censor. The models were fit separately to British and Italian samples.

3.2.1. Cognitive Empathy

The first hypothesis of this study was that cognitive empathy predicts street harassment tolerance. Positive empathy was skewed towards the lower values, and negative empathy was normally distributed around the middle rating (Appendix A, Figure A1). Cognitive empathy (composite score) was negatively associated with street harassment tolerance in both samples (Spearman’s r(246) = −0.55, p < 0.001). In further exploration (Appendix A,

Table A2. Relationship of cognitive empathy to street harassment tolerance (standardized parameter estimates, Tobit regression).

Cognitive Empathy	Before Video	After Video
British
Positive emotion	0.68	0.66
Negative emotion	−0.47	−0.45
Composite score	−0.46	−0.45
Italian
Positive emotion	0.66	0.76
Negative emotion	−0.60	−0.72
Composite score	−0.43	−0.52

), greater assumptions of positive emotion, lesser assumptions of negative emotion, and composite scores were used to predict street harassment tolerance. The associations between composite empathy scores and street harassment tolerance are visualised in Figure A2. Model comparison using AIC demonstrated that composite scores provided the best fit to data, so only composite scores were taken forward to a combined model with SDO (Appendix A,

Table A3. Regression tables, predicting street harassment tolerance from SDO and cognitive empathy composite score.

Before Video
British Sample
Variables	Model R2	b	SE (b)	z	p
	0.43
(Constant)		32.0	5.07	6.3	<0.001
Cognitive empathy		−0.38	0.07	−5.84	<0.001
Social dominance orientation (SDO)		0.26	0.04	6.08	<0.001
Italian Sample
Variables	Model R2	b	SE (b)	z	p
	0.34
(Constant)		31.3	6.1	5.1	<0.001
Cognitive empathy		−0.38	0.07	−5.5	<0.001
Social dominance orientation (SDO)		0.24	0.06	4.18	<0.001
After Video
British Sample
Variables	Model R2	b	SE (b)	z	p
	0.37
(Constant)		30.0	5.5	5.4	<0.001
Cognitive empathy		−0.38	0.07	−5.3	<0.001
Social dominance orientation (SDO)		0.26	0.05	5.4	<0.001
Italian Sample
Variables	Model R2	b	SE (b)	z	p
	0.34
(Constant)		39.0	6.3	6.2	<0.001
Cognitive empathy		−0.47	0.07	−6.6	<0.001
Social dominance orientation (SDO)		0.20	0.06	3.4	0.001

). All effects were replicated across British and Italian samples.

3.2.2. Social Dominance Orientation

The second hypothesis was that a higher SDO (Appendix A, Figure A3) would predict higher street harassment tolerance. A higher SDO predicted greater street harassment tolerance both pre- and post-video in both samples (British pre-video: b = 0.31, post-video: b = 0.30; Italian pre-video: b = 0.31, post-video: b = 0.27; all ps < 0.001). These results were replicated across both cultural contexts (Appendix A, Figure A4).

3.3. Combined Model

When including both SDO and cognitive empathy in the same model, both predictors independently contributed to street harassment tolerance (Appendix A, Table A3). Interaction effects with nationality were not significant (ps > 0.45), suggesting that the associations between SDO, cognitive empathy, and street harassment tolerance did not differ between the British and Italian samples. Age was positively associated with street harassment tolerance (b = 0.16, p = 0.001). Given the significant age difference between British and Italian participants (t(242) = 6.5, p < 0.001), a covariate of age was included in this analysis. Education and social relationship histories were not significantly associated with street harassment tolerance.

3.4. Population Comparison

The final aim of this study was to examine whether street harassment tolerance, cognitive empathy, and SDO scores would differ across the British and Italian cultures. To assess these differences, independent sample t-tests were conducted comparing British and Italian participants across the three scales. The results indicated that British participants reported significantly higher street harassment tolerance compared with Italians (M_diff = 2.12, 95% CI [−4.09, −0.15], t(241.7) = −2.12, p = 0.03). However, this difference reflected a small effect size (r = 0.14), according to Cohen’s (1988, 1992) benchmarks. Significant differences were also observed in cognitive empathy scores (M_diff = −8.14, 95% CI [−11.24, −5.04], t(246.12) = 5.17, p < 0.001), with British participants displaying lower cognitive empathy than their Italian counterparts. This difference corresponded to a medium effect size (r = 0.32). Additionally, SDO scores were significantly higher among British participants compared with Italians (M_diff = 6.66, 95% CI [4.89, 8.44], t(246.29) = 7.39, p < 0.001), with a medium effect size (r = 0.43).

4. Discussion

This study aimed to investigate the relationship between cognitive empathy, SDO, and street harassment tolerance in British and Italian heterosexual male populations. Beliefs in relatively more positive and fewer negative emotions in the victim, as well as preferences for social hierarchies and inequality, were associated with greater street harassment tolerance. These effects were consistent across two distinct cultures.

This study examined men’s cognitive empathy towards a female victim’s emotional experience during video-recorded street harassment episodes. The study’s average cognitive empathy scores suggested that men generally understood street harassment as a negative experience for the recipient, corresponding to a low tolerance for such behaviour. However, within this range, lower cognitive empathy for the victim was linked to greater tolerance for street harassment. This aligned with recent findings suggesting that men who self-report harassment may mistakenly believe they are expressing affection to evoke a positive response (DelGreco et al., 2021) and with earlier theories proposing that men often perceive sexual harassment as less harmful and more welcome than it actually is (Jacques-Tiura et al., 2007; Stockdale, 1993).

The positive relationship between SDO and tolerance for street harassment could arise from multiple mechanisms. SDO reflects a desire for dominance of one’s group over others. Early research suggested that men may appraise women’s distress following street harassment as a means to establish power (di Leonardo, 1981). This was based on the belief that eliciting a stress response increased the perpetrator’s sense of power. Street harassment can thus be seen as an attempt to exert control over another person, regardless of whether it is expected to result in a positive (DelGreco et al., 2021) or negative (di Leonardo, 1981) experience for the victim. Additionally, street harassment may be motivated by assumptions that women appreciate the attention, particularly if the perpetrator feels powerful in society (Kunstman & Maner, 2011), by the dehumanisation of outgroups (Ho et al., 2015), or by the reinforcement of patriarchal social scripts (Seal & Ehrhardt, 2003). While the link between SDO and street harassment tolerance has been established, further research is needed to understand this relationship better.

While cognitive empathy and SDO independently relate to street harassment tolerance, the two measures were also correlated with each other, consistent with the prior literature (Duckitt & Sibley, 2010; McFarland, 2010; Sidanius et al., 2013). Interestingly, SDO was correlated with expectations of positive experiences for the victim, but not negative ones, supporting the link between power-related motivations and assumptions of a positive response to the harasser’s attention.

Given the association between cognitive empathy and street harassment tolerance, interventions targeting empathy enhancement may effectively reduce acceptance of harassment. Prior research on counteracting hate speech (Sportelli et al., 2025) has demonstrated that fostering empathy through digital intergroup contact and targeted training can reduce tolerance for harmful social behaviours. Applying similar interventions to street harassment tolerance, particularly among younger populations, could be a promising avenue for prevention. Digital education programs, interactive perspective-taking exercises, and structured empathy training may be beneficial.

Similarly, SDO-focused interventions could benefit from established approaches in moral education. Character education programs that emphasise prosocial values, ethical reasoning, and social justice awareness (Brown et al., 2023) have effectively reduced hierarchical and dominance-driven attitudes. Moreover, innovative interventions that highlight the emotional risks of social dominance and the value of humility (D’Errico, 2019) could further encourage attitudinal shifts. By emphasising the personal and social costs of maintaining rigid dominance hierarchies, such programs may contribute to lowering tolerance for street harassment.

Despite its limited impact on overall attitudes, the experimental effect of the video intervention can be acknowledged. While the video significantly reduced street harassment tolerance in British participants, no such effect was observed in the Italian sample. This discrepancy may reflect cultural differences in how video-based interventions are perceived or processed. Additionally, language differences may have played a role as subtitled content may not have the same impact as native-language material, potentially affecting engagement or comprehension. More generally, the findings aligned with prior research indicating that video media alone has a limited effect on harassment attitudes (Darnell & Cook, 2009). A potential floor effect may also explain the results as the baseline tolerance was already low in both samples. This underscores the challenge of shifting deeply ingrained attitudes and suggests that more immersive, interactive approaches may be necessary to foster meaningful change. Notably, the associations between cognitive empathy, SDO, and street harassment tolerance remained stable pre- and post-intervention, indicating that these psychological constructs may influence one another bidirectionally and persist independently of brief interventions.

Associations between cognitive empathy, SDO, and street harassment tolerance were consistent across British and Italian samples, yet the significant age difference between the groups warrants further consideration. Generational shifts in attitudes towards gender equality and evolving societal norms may have shaped participants’ views on street harassment. The younger Italian sample exhibited lower tolerance and greater cognitive empathy, suggesting that younger individuals may be more receptive to changing social narratives around harassment. Conversely, older generations may hold more tolerant views, which could be associated with baseline attitudes and interventions’ effectiveness.

This age-related trend highlights the importance of targeting educational efforts towards adolescents and young adults. Reinforcing empathy-building strategies early in social development may foster long-term attitudinal shifts, making interventions more impactful. Despite some cultural differences, our findings mostly overlapped, suggesting shared opportunities for preventative education in both countries—specifically, promoting awareness of the victim’s negative experience and challenging patriarchal social hierarchies in public spaces.

Limitations and Future Directions

As with all studies, these experiments had limitations that future research should address. First, participants were recruited opportunistically through an online platform and social media, which may have introduced selection bias (Smart, 1966). Although this method provided a heterogeneous sample in terms of age and education, it may have led to an overrepresentation of certain demographic groups who are more active or engaged in these platforms, potentially skewing the sample in ways that affect the attitudes measured. For instance, individuals who participate in online platforms or social media may already possess a certain level of social awareness or engagement with social issues, which could influence their attitudes towards street harassment and tolerance of social inequalities. Future research should explore how recruitment strategies influence attitudes, especially by comparing different recruitment methods (e.g., online vs. in person) and examining their impacts on the study outcomes.

Second, the study relied on self-report measures, which can lead to underreporting socially undesirable attitudes (Jobe, 2003; Schwarz & Oyserman, 2001), possibly contributing to the floor effect observed in tolerance measures. While the anonymity of the online format likely mitigated this risk, alternative methods could further reduce it.

Third, the study focused on nationality but did not assess whether participants lived most of their lives in their native countries, limiting the ability to draw clear conclusions about cultural differences. Additionally, the correlational nature of the study prevented causal inferences. We could not conclusively determine whether lower empathy or higher SDO caused increased street harassment tolerance or vice versa. Intervention studies, such as those by Diehl et al. (2014), which showed that increasing perspective-taking reduced the likelihood of engaging in sexual harassment, are recommended as the next steps.

Finally, the study explored two Western populations, so further research is needed to understand if these attitudinal and cognitive relationships hold across more diverse cultural contexts, for example, in Japan, where patriarchal norms and sociocultural pressures to maintain harmony have historically influenced attitudes towards sexual harassment differently from those in Western societies (Xilun Pang & Tomlinson, 2023). Moreover, our study focused on heterosexual men, which limited the generalizability of our findings to other types of street harassment. Specifically, the findings may not extend to non-heterosexual men. Future research should explore the role of cognitive empathy and SDO in street harassment tolerance across a broader range of sexual orientations, gender identities, and cultural backgrounds. This would provide a more comprehensive understanding of how these psychological factors interact with harassment tolerance across diverse groups. Additionally, future studies should examine how different forms of harassment, including those targeting LGBTQ+ individuals and racial minorities, may be associated with empathy and dominance motivations (Fileborn & O’Neill, 2023).

5. Conclusions

Since the responsibility lies with perpetrators and their peers to reduce harmful street harassment towards women, it is essential to understand the underlying attitudes and cognitions. This study provides initial insight into the independent roles of cognitive empathy and SDO in street harassment tolerance across two distinct cultures, identifying potential targets for positive social change. Addressing perpetrators’ perspectives holds significant potential for preventing street harassment at its source rather than merely mitigating its effects. While the link between tolerance of sexual harassment and street harassment is modest (DelGreco et al., 2021), changing empathy levels and attitudes towards power could have an impact through reduced peer support for perpetrators, policy changes, cultural shifts, and increased allyship for victims. By implementing these strategies early, particularly among younger populations, it may be possible to foster lasting societal change in attitudes towards street harassment.

Our findings suggest that such interventions may be effective across multiple national contexts.