Originally coined by activist Tarana Burke in 2006, #MeToo gained momentum in 2017 following a viral tweet by Alyssa Milano encouraging women to use the hashtag to share their stories of sexual harassment and assault (Williamson et al., 2020). The #MeToo movement highlighted the lifetime prevalence of sexual harassment toward women. Millions of women posted about their experiences under the hashtag in the first 48 h, catalysing research assessing the impact sexual harassment has on women’s lives (Chawla et al., 2021; Keplinger et al., 2019). The academic literature matches personal accounts by providing correlational evidence that experiences of sexual harassment have detrimental effects on women’s well-being (Chan et al., 2008; Sojo et al., 2016; Willness et al., 2007).

Though informative, critical shortcomings of the extant literature limit conclusions about the impact of sexual harassment on women. First, research has predominantly examined sexual harassment within the workplace, limiting generalizability to the many other contexts in which sexual harassment occurs. Second, prior studies have not compared those who reported experiencing sexual harassment to a matched control group rendering it unclear whether the lower well-being associated with sexual harassment is the result of selection effects or key confounding variables. Third, research has almost exclusively focused on well-being outcomes, with no prior examination of the broader socio-political effects of experiencing sexual harassment on women’s perceptions of societal equality or support for the gender hierarchy. This oversight is crucial for understanding how significant experiences that often emerge due to gender hierarchies of power may produce changes in attitudes that have broad implications for gender inequality and social change.

In the current study, we overcome these key limitations by drawing on national probability survey data and utilizing propensity score matching to identify (1) women who reported sexual harassment in the past year and (2) a matched control group of women who were similar in outcome and demographic variables in the previous year but did not report sexual harassment (Nmatched pairs = 609). Advancing prior examinations of the impact of sexual harassment of well-being, we compared pre- and post-event measures of well-being to test whether women who reported sexual harassment experienced declines in well-being, and whether these changes in well-being were significantly different to the matched control group. Extending the outcomes in prior investigations, we also compared pre- and post-event levels of system justification, including the perceived fairness of gender relations and society in general. Our aim was to advance understanding of the broader socio-political repercussions of sexual harassment by testing whether experiencing sexual harassment not only impacts women’s well-being, but also affects their perceptions of gender relations and society in ways that have important implications for social change.

Sexual Harassment: Definitions, Prevalence, and Well-Being Impact

Sexual harassment generally entails (1) unwelcomed and unwanted behaviour that (2) is offensive, humiliating, and/or intimidating to the recipient, (3) regardless of the intent of the perpetrator (Ministry for Women, 2019; Quick & McFadyen, 2017). Sexual harassment can vary in form (e.g., sexual nature or gender-based discrimination, verbal or physical, in-person or online, once or repeatedly over time) and the domain in which it occurs (e.g., employment, education, housing; Fitzgerald & Cortina, 2018; McDonald, 2012). Examples include offensive sexual remarks or jokes, inappropriate touching, and preferential treatment for compliance or threats for non-compliance to sexual or romantic advances (Fitzgerald & Cortina, 2018; McDonald, 2012).

Sexual harassment is shockingly prevalent. National data of over 12,000 people in 2020 found that 84% of women in the UK experienced sexual harassment at least once during their lifetime (2020 Sexual Harassment Survey; Adams et al., 2020). A nation-wide US-based study similarly found that 81% of women experienced sexual harassment once or more in their lifetime (Kearl et al., 2019). In contrast, national data surveying 42,000 people from 28 countries across the European Union found that 45–55% of women reported experiencing sexual harassment of any type since aged 15 (European Union Agency for Fundamental Rights, 2014). No similar national-level studies have been conducted in New Zealand, where the current study is situated. However, the New Zealand Crime and Victims Survey (NZCVS; Ministry of Justice, 2022) provides prevalence data for sexual violence, which includes more serious forms of sexual harassment such as unwanted sexual touching. Pooled data of roughly 30,000 New Zealanders (2017–2021) indicate that 35% of women experienced sexual assault at least once in their lifetime.

Prior research has primarily focused on the job-related, physical, and psychological consequences of sexual harassment for women (e.g., Feldblum & Lipnic, 2016; Fitzgerald et al., 1997; Munson et al., 2000). Three recent meta-analyses illustrate that sexual harassment correlates with (a) poorer job-related outcomes (e.g., lower job satisfaction and commitment, and increased work withdrawal; sample-size-weighted mean rs = 0.14–0.32), (b) poorer physical health (i.e., greater health symptoms like headaches, exhaustion; rs = 0.17–0.29), and (c) poorer psychological well-being (rs 0.12–0.30), including greater depressive, anxiety and PTSD symptoms, increased psychological distress, and lower self-esteem and life satisfaction (Chan et al., 2008; Sojo et al., 2016; Willness et al., 2007).

Despite providing evidence for the negative impact of sexual harassment, meta-analyses highlight important methodological shortcomings that narrow the conclusions that can be drawn about the effects of sexual harassment on women. First, most studies use samples of women in (specific) workplaces, limiting generalizability of the results to the myriad contexts in which sexual harassment occurs. For example, national data from the UK suggest that only a third (31%) of sexual harassment experiences occurred at the workplace (Adams et al., 2020). Although workplace experiences of sexual harassment are associated with poorer well-being and job-related outcomes, we broaden the scope by examining the effects of any instance of sexual harassment inside or outside the workplace.

Second, prior studies have not compared women who reported experiencing sexual harassment to those of a matched control group who did not report experiencing sexual harassment. This oversight limits inferences regarding the impact of sexual harassment on well-being because the two groups may vary in important ways beyond the experience of sexual harassment. Propensity score matching addresses this limitation by approximating the conditions of a randomized controlled experiment, thereby strengthening causal inferences (Rosenbaum & Rubin, 1983; Thoemmes & Kim, 2011). The matching process involves identifying a control group with comparable demographics and outcome levels but who did not experience the event (Austin, 2011; Foster, 2010; Stuart, 2010). Recent studies employing propensity score matching have found that some documented changes associated with important events (e.g., divorce, major illness, retirement) may not be due solely to the focal event as the changes also naturally occurred in a matched control group, highlighting the importance of accounting for confounding factors (Buecker et al., 2021; Costanzo et al., 2009; van Scheppingen & Leopold, 2020).

Given the absence of propensity matched control groups in prior studies assessing the effects of sexual harassment, the observed associations between reported sexual harassment and well-being may be, at least in part, explained by pre-existing differences in well-being and other relevant factors that shape the likelihood of experiencing sexual harassment. For example, women may deny gender discrimination to sustain their own well-being, meaning that women with lower well-being may be more aware of experiencing sexual harassment and, thus, more likely to report it (see Napier et al., 2020). Moreover, the prevalence of sexual harassment is higher among younger women (under 35 years old; Adams et al., 2020), and younger people also exhibit stronger drops in well-being when experiencing workplace sexual harassment (see Chan et al., 2008). By not accounting for these likely selection effects and other important confounding factors, prior results may have provided an inaccurate picture of the effects of sexual harassment on well-being. In the current research, we address these possible selection effects by using propensity score matching to identify a matched control group that is similar to women who reported sexual harassment in both pre-existing levels of well-being and other important demographic variables (e.g., age, religiosity) but who did not report sexual harassment (Nmatched pairs = 609). We then compare pre- and post-event differences in well-being across these two equally sized groups with comparable pre-event well-being and demographic profiles to provide a stronger test of the impact of sexual harassment on women’s well-being.

Broader Implications of Sexual Harassment: Socio-Political Impact

Another important limitation of prior research on the effects of sexual harassment is that this work predominantly focuses on women’s well-being and mental health (e.g., Chan et al., 2008). As a gendered behaviour that often emerges from, or in protection of, gender hierarchies of power (e.g., Begany & Milburn, 2002), sexual harassment is a common experience that could lead women to resign to—or even justify—the status quo of gender inequality. Alternatively, these experiences could fundamentally alter women’s perceptions of society, particularly regarding gender relations. Although some work notes that the sexual objectification of women can undermine their sense of political agency (see Calogero, 2013a, b, 2017; Calogero & Tylka, 2014), no prior empirical investigations have considered the specific effects of sexual harassment on women’s societal perceptions or beliefs about gender relations. This oversight is surprising given the myriad movements, including #MeToo, that aim to promote social change by highlighting the prevalence—and challenging the norms—of sexual harassment (see Chawla et al., 2021). This oversight is also theoretically important because examining changes in women’s societal perceptions of gender relations will advance understanding about whether key gendered experiences, such as sexual harassment, impact women’s attitudes in ways that support or challenge the gender hierarchy.

Women’s perceptions of society and gender relations have important implications for gender inequality and social change. An important line of theoretical and empirical work demonstrates that people’s perceptions of society may often serve a system-justifying function (Jost, 2020; Jost & Kay, 2005). Applying system justification theory to gender relations, women may undermine their self-interest and support the unfair and unequal gender hierarchy because it protects them from the negative well-being outcomes of discrimination (see Bahamondes et al., 2019; Becker & Wright, 2011; Napier et al., 2020). For example, women may endorse system-justifying beliefs, including perceiving gender relations as fair and just, to gain a sense of certainty, particularly if they are high in epistemic needs (Howard, Sibley et al., 2022). Women are also more likely to deny sexism is a problem and endorse traditional gender stereotypes when primed with the idea that merit and hard work underlie people’s progress in life (McCoy & Major, 2007). Exposure to benevolent stereotypes of women and men can also increase women’s system justification, including their support for the unequal gender system (Jost & Kay, 2005).

These palliative effects of system justification are important because they ultimately undermine women’s resistance to gender inequality (see Bahamondes et al., 2019; Becker & Wright, 2011; Calogero, 2013b, 2017; Calogero & Tylka, 2014; Napier et al., 2020). Applying system-justifying principles to the context of sexual harassment, experiences with sexual harassment may fail to change—or perhaps even increase—support for the unequal status quo, including the perception that the (gendered) system is fair (see Calogero, 2013a; Calogero & Tylka, 2014; Jost et al., 2003). Yet surprisingly, no prior research has tested whether the significant gender-related experience of sexual harassment affects women’s perceptions of society or their support (vs. challenge) of gender inequality. Moreover, prior studies that reveal system-justifying effects have primarily occurred outside the context of specific, harmful injustices or events that make the inequalities and dangers of the gender hierarchy apparent (see Osborne et al., 2019). When such experiences conflict with or override the well-being benefits of system justification, women may be motivated to change the social system (see Jost et al., 2001). Providing some preliminary evidence, Becker and Wright (2011) found that women exposed to hostile sexism (i.e., hostility toward women who threaten the gender hierarchy) perceived gender relations as less fair.

In sum, experiences of sexual harassment could either lead women to justify the status quo of gender inequality or undermine their perceptions that gender relations are fair. Given sexual harassment is inherently hostile and reinforces men’s dominance, we expected that the costs associated with sexual harassment would outweigh any potential benefits of supporting an unfair social system. Thus, we tested—for the first time—whether experiences of sexual harassment were associated with a reduction in women’s perceptions that gender relations and society are fair by comparing pre- and post-event differences in system justification between women who experienced sexual harassment and a matched control group who had comparable levels of pre-event system justification. We focused primarily on gender-specific system justification involving the perception that gender relations in society are fair. Although we expected the gendered nature of sexual harassment to impact women’s perceptions of gender relations most strongly, we also included a broader measure of general system justification to examine if the impact of sexual harassment generalizes to more global perceptions of fairness in society. The results have important implications for understanding whether discriminatory experiences in specific domains promote change in domain-specific beliefs as well as broader beliefs associated with support for social change.

Present Study

In the current study, we aim to provide stronger evidence of the impact of sexual harassment on women’s well-being by examining pre-to-post-event changes in well-being (life satisfaction, psychological distress) across women who reported sexual harassment in the past year and matched controls of women who did not report sexual harassment (Nmatched pairs = 609). Additionally, we expand the almost exclusive prior focus on well-being to assess the broader impact of sexual harassment on (gender) system justification. Sexual harassment of women is often a direct consequence of men’s attempts to protect gender hierarchies of power (Begany & Milburn, 2002), and, therefore, likely influences women’s perceptions of gender relations and society in ways that either justify or challenge the status quo of (gender) inequality. Thus, we also assessed whether, compared to matched controls, women who reported sexual harassment were less likely to perceive gender relations as fair and just (gender-specific system justification). We also assess women’s broader beliefs about the fairness of society in general (general system justification) to examine if the impact of sexual harassment generalizes to more global societal perceptions.

As with any potentially traumatic experience, it would be unethical to subject women to experience sexual harassment using an experimental design. Because random assignment to infer causality is not feasible in these cases (Foster, 2010), propensity score matching provides an alternative approach towards examining the impact that sexual harassment has on women’s well-being and system justification. However, the degree to which propensity score matching strengthens causal inferences is dependent on the selection of appropriate co-variates that predict the outcomes and the likelihood of experiencing the event (Foster, 2010). To minimize any potential bias in acknowledging and reporting experiences of sexual harassment, we first matched participants on pre-event levels of the well-being and system justification outcomes (see Bahamondes et al., 2021; Napier et al., 2020). For example, women lower in well-being or system justification may be more likely to report sexual harassment (see Napier et al., 2020). Thus, any between-group differences following the event may be explained by lower baseline levels among women who reported sexual harassment. However, matching women who reported sexual harassment on pre-event well-being and system justification in the prior year eliminates the possibility that any post-event differences are simply due to baseline levels of well-being and system justification.

We also matched participants on several important demographic variables that are relevant to well-being, system justification, and sexual harassment (see Method section for more information). To illustrate, we matched participants on age, given that younger people not only express greater general well-being (Steptoe et al., 2015) but also report higher rates of sexual harassment than their older counterparts (Adams et al., 2020). Again, creating a control group that is matched on these key covariates provides stronger evidence that any pre-to-post-event differences are due to the experience of sexual harassment rather than pre-existing factors that may make women more susceptible to sexual harassment, poorer well-being, and perceptions that gender relations and society are unfair. Altogether, our application of propensity score matching to assess pre- and post-event changes in well-being and system justification strengthens the evidence for the effect of women’s experiences of sexual harassment on important outcomes.

Method

Participants and Sampling Procedure

We utilized data from Time 9 (2017) to Time 14 (2022) of the New Zealand Attitudes and Values Study (NZAVS), which incorporates the first four waves using a life events measure to assess the experience of sexual harassment (Time 10–14) and pre-event data from the wave immediately prior to reports of sexual harassment (Time 9–13). Participants were sampled from the electoral roll and closely represented the general New Zealand population, with a few biases: women and Europeans are overrepresented by roughly 10%, whereas Asian people are underrepresented by roughly 5%. See Sibley (2021) for details and full materials regarding the NZAVS. The NZAVS was reviewed and approved by The University of Auckland Human Participants Ethics Committee (Reference Number: 014889).

To capture the largest sample of women who spontaneously reported sexual harassment in the past year and account for selection effects, we used pre-event data to identify a control group. Across the four timepoints measuring sexual harassment (Time 10–14), we identified the first instance in which women (a) reported sexual harassment and (b) also completed the prior wave of the survey and thus provided pre-event data. For example, if a participant reported sexual harassment at Time 11, we used their responses at Time 10 as pre-event data. A total of 670 participants met these inclusion criteria. However, to ensure optimal matching, we excluded women who had missing data across outcome (n = 21) and demographic (n = 38) variables used in the matching procedure. The potential sample of women to draw the matched control group included women who did not report sexual harassment across any of the four waves and had two consecutive waves of available data. To create analogous pre-post data as the event group, we took the first two waves of available data for the potential control participants. Overall, the matching process included 611 women who reported sexual harassment and provided pre-event data, and 25,197 potential controls with analogous pre-post data.

Propensity Score Matching

Using the one-to-one propensity score matching function in SPSS 29 (see Thoemmes, 2012), we matched women who reported sexual harassment with women who had a similar propensity score but did not report sexual harassment (see Cross et al., 2021; Sibley et al., 2020). Matching participants who reported sexual harassment with similar women who did not report sexual harassment increases confidence that any observed differences between the two groups are likely due to the experience of sexual harassment rather than other factors (e.g., selection effects or maturation processes; see Thoemmes & Kim, 2011). To address selection effects in the reporting of sexual harassment (see Bahamondes et al., 2021; Napier et al., 2020), we matched on the pre-event measures of well-being and system justification (life satisfaction, psychological distress, gender-specific system justification, general system justification). We also matched on a range of demographic variables that may influence both well-being and system justification, and the likelihood of experiencing sexual harassment: age, ethnicity (Māori, Pacific, Asian), education level, having a partner, being a parent, employment status, religiosity, born in New Zealand, and timepoint the pre-event data were drawn. We prioritized exact matches and used a match tolerance of 0.01, which restricts the distance between propensity scores for two participants to be matched to 0.01 SDs of the logit of the propensity score (Austin, 2011). The matching process yielded 609 matches (i.e., there were only two failures to match).

Table 1 first provides the descriptive statistics for the whole sample of potential controls followed by the group of women who reported sexual harassment. Women who reported sexual harassment clearly have lower levels of well-being and system justification than women who did not report sexual harassment (ts > 7.53, ps < 0.001 across all pre-event outcomes) highlighting selection effects involving women lower in well-being and system justification being more likely to report sexual harassment (see Bahamondes et al., 2021; Napier et al., 2020). Crucially, however, the matching process identified a control group of women with the same lower levels of pre-event well-being and system justification. The event group of women who reported sexual harassment and the matched control group (Nmatched pairs = 609) did not differ across any of the pre-event outcomes (see final columns for test of mean differences across groups), thereby eliminating the possibility that any post-event effects are the result of pre-existing differences or selection effects (see Bahamondes et al., 2021; Napier et al., 2020). Additionally, the matching process identified a control group with no differences in pre-existing demographic differences with two exceptions. The group who reported sexual harassment remained slightly younger than the matched control group, but notably the difference in age was substantially reduced compared to the whole sample of potential controls (see Table 1). Data from the group who reported sexual harassment also were more likely to be drawn from earlier time waves as a function of taking the first two consecutive waves of available data. In sum, we identified a control group who had similarly low levels of well-being and system justification, and a similar demographic baseline, but did not report sexual harassment. Thus, the matched control group provides a critical baseline to evaluate whether experiencing sexual harassment predicts unique pre-to-post-event changes.

Table 1 Descriptive statistics for variables used in the matching procedure
Full size table

Measures

Propensity Matching Variables

Coded responses to open-ended questions were used to assess the variables partner (“What is your relationship status?”; 0 = no partner, 1 = partner), parent (“How many children have you given birth to, fathered, or adopted?”; 0 = not a parent, 1 = parent), education (“What is your highest level of qualification?”; 0 = low to 10 = high; New Zealand Qualifications Authority, 2012), and born in New Zealand (“Where were you born?”; 0 = outside New Zealand, 1 = New Zealand). A yes (1) or no (0) question measured the variables employed (“Are you currently employed?”) and religious (“Do you identify with a religion and/or spiritual group?”). Ethnicity (Māori, Pacific, and Asian: 0 = no, 1 = yes) was assessed using a checklist and coded open-ended question asking, “Which ethnic group(s) do you belong to?” (Stats, 2020). One question asking, “What is your date of birth?”, was used to calculate participants’ age. The mean of the wave prior to reporting sexual harassment (Time 9 to Time 13) was used to capture timepoint drawn from.

Sexual Harassment

Sexual harassment was assessed using the Broad Inventory of Specific Life Events (BISLE; Howard et al., 2022). The BISLE involves a checklist of 15 common life events followed by an open-ended question asking, “Finally, have you experienced any significant life events in the past year?” Sexual harassment is one of the 15 checklist probes and is also an event coded from open-ended responses. Participants are asked to indicate whether “someone sexually harassed you” within the past year. Responses are categorized into yes (1) or no (0). Because the open-ended question asks participants to report other events not covered on the checklist, we expected that most reports of sexual harassment would be produced from the checklist. For example, we found that 99% of women who reported sexual harassment at Time 10 did so using the checklist. Notably, the BISLE was included on the last page of the questionnaire, and the well-being and system justification items were interspersed with other Likert items throughout the earlier pages. Thus, it is unlikely that being asked about sexual harassment in the final component of the questionnaire primed responses to the well-being and system justification outcomes assessed earlier.

Outcomes

All outcomes were measured on 1 (strongly disagree) to 7 (strongly agree) scale, except for the psychological distress measure, which was rated on a 0 (none of the time) to 4 (all of the time) scale.

Life satisfaction was measured using the average (i.e., scale mean) of two items from the Satisfaction with Life Scale (Diener et al., 1985): “I am satisfied with my life” and “In most ways my life is close to ideal” (α = 0.79). Psychological distress was measured using the Kessler-6 screening scale (Kessler et al., 2010), in which participants are asked “During the last 30 days, how often did…” followed by six items that were averaged (e.g., “…you feel hopeless?”, “…you feel restless or fidgety?”, “…you feel worthless”; α = 0.86).

Gender-specific system justification was measured using the average of two items adapted from Jost and Kay (2005): “In general, relations between men and women in New Zealand are fair” and “Men and women both have a fair shot at wealth and happiness in NZ” (α = 0.68). General system justification was measured using the average of four items from Kay and Jost (2003): “In general, the New Zealand political system operates as it should”, “In general, I find New Zealand society to be fair”, “Everyone has a fair shot at wealth and happiness in New Zealand”, and “Most of New Zealand’s policies serve the greater good” (α = 0.63).

Results

To test whether reported experiences of sexual harassment predicted differences in well-being and system justification compared to a matched control group, we conducted a series of 2 × 2 Mixed Model ANOVAs with event group (reported sexual harassment vs. matched control) as a between-person factor and timepoint (pre- vs. post-event) as a within-person factor. This analytic strategy allowed us to test whether well-being and system justification significantly differed after the experience of sexual harassment compared to a control group matched on these outcomes (and other key demographics) prior to the event. We expected a significant Event × Timepoint interaction to reveal no differences between groups prior to the event, but significant differences in well-being and system justification after the event. Follow-up pairwise comparisons (with Bonferroni adjustments) examined within-group differences in outcomes prior to the event (pre-event) and after the event (post-event).

As shown in Table 2, the Event × Timepoint interaction was significant for all four outcomes: life satisfaction, psychological distress, gender-specific system justification, and general system justification. Table 3 displays the mean levels of well-being and system justification for women who reported sexual harassment and the matched controls along with pairwise comparisons between groups. As reported above, across all four outcomes, the groups did not significantly differ prior to the event. After the event, however, women who reported sexual harassment had significantly lower life satisfaction, greater psychological distress, lower gender-specific system justification, and lower general system justification compared to matched controls.

Table 2 Results for 2 (event: reported sexual harassment vs. matched control) × 2 (timepoint: pre- vs. post-event) ANOVAs predicting well-being and system justification
Full size table
Table 3 Pre- and post-event descriptive statistics and mean differences across event groups
Full size table

Although these results reveal differences across groups after the event, we also wanted to test whether those who reported sexual harassment experienced a change in well-being and system justification from their pre-event levels. Thus, we conducted pairwise comparisons to test whether pre-post changes occurred within groups. Focusing first on the well-being outcomes, women who reported sexual harassment exhibited significant pre-to-post decreases in life satisfaction (Mdiff = -0.14, 95% CI [-0.22, -0.05], p = .002, d = 0.09) and increases in psychological distress (Mdiff = 0.11, 95% CI [0.06, 0.16], p < .001, d = 0.13). These within-person changes contrasted to the matched control group who showed smaller effects in the opposite direction, including significant increases in life satisfaction (Mdiff = 0.09, 95% CI [0.01, 0.18], p = .034, d = 0.06) and decreases in psychological distress (Mdiff = -0.10, 95% CI [-0.15, -0.05], p < .001, d = 0.12).

In terms of system justification variables, women who reported sexual harassment exhibited significant pre-to-post decreases in gender-specific system justification (Mdiff = -0.15, 95% CI [-0.06, 0.25], p = .001, d = 0.10) and smaller but still significant decreases in general system justification (Mdiff = -0.09, 95% CI [-0.17, -0.02], p = .013, d = 0.08). These within-person changes contrasted to the matched control group who showed increases in gender-specific system justification (Mdiff = 0.10, 95% CI [0.00, 0.19], p = .043, d = 0.07) and no changes in general system justification (Mdiff = 0.06, 95% CI [-0.02, 0.13], p = .130, d = 0.04). In sum, women who reported sexual harassment experienced significant reductions in well-being and perceptions that gender relations and society in general are fair, and these changes did not emerge among women who were comparable on outcomes and demographics prior to the event but did not report sexual harassment.

Discussion

Sexual harassment harms women’s well-being, but does it change women’s perceptions of gender relations and society? No prior investigations have tested whether this important gendered experience affects women’s perceptions of gender relations and society in ways that support or challenge the gender hierarchy and social inequity. Utilizing propensity score matching, we identified (1) women who reported sexual harassment and (2) a matched control group who had similar pre-event outcomes levels and demographic profiles but did not report sexual harassment (Nmatched pairs = 609). By examining pre-to-post differences across groups, we provided stronger evidence of the impact of sexual harassment on well-being and new evidence demonstrating the impact of sexual harassment on women’s system justification. Women who reported sexual harassment experienced significant pre-to-post declines in well-being (lower life satisfaction, higher psychological distress) and significant pre-to-post reductions in perceptions that gender relations, and broader society, are fair. Critically, these changes were significantly different than matched controls. These novel results suggest that experiences of sexual harassment not only negatively impact well-being, but also reduce women’s support for the gender hierarchy and general status quo.

Theoretical Implications: Broader Impacts of Sexual Harassment on Women

Studies investigating the impact of life events on well-being are increasingly using more rigorous methods, such as propensity score matching, to strengthen causal inferences (e.g., van Scheppingen & Leopold, 2020). However, propensity score matching has not been used to examine women’s experiences with sexual harassment. Accordingly, by not accounting for likely selection effects and other important confounding factors, prior studies may provide an inaccurate picture of the effects of sexual harassment on well-being. We addressed this limitation by employing propensity score matching to identify a matched control group that was similar in both their pre-existing levels of well-being and other important demographic variables (e.g., age, religiosity). Our results illustrated that women who experienced sexual harassment demonstrated significant declines in well-being that were not evident in matched controls. Instead, matched controls reported smaller but significant improvements in well-being in the following year. Such increases may reflect either regression to the mean given lower starting levels of well-being than the general sample (see Table 1) or a typical pattern of recovery from drops in well-being (see Bonanno, 2004; Bonanno et al., 2011). Regardless, the declines in well-being experienced by women who reported sexual harassment reflect a significant departure from the baseline changes shown by matched controls.

Beyond providing stronger evidence that sexual harassment impacts well-being, the current study uniquely illustrates broader socio-political consequences, which has important theoretical implications. Inspired by system justification theory, we investigated whether reported experiences of sexual harassment would outweigh women’s motivation to justify an unfair gendered system (see Jost et al., 2001; Osborne et al., 2019). Prior work demonstrates that women may continue to support an unfair and unequal hierarchy to protect against the negative well-being outcomes of discrimination (see Bahamondes et al., 2019; Becker & Wright, 2011; Jost & Kay, 2005; Napier et al., 2020). Such system-justifying effects are critical to understand because they contribute to the maintenance of gender inequality. Yet, the current study provides important new evidence that women who report experiences with sexual harassment are more likely to reject the unfair gender hierarchy: compared to matched controls, women who reported sexual harassment showed significant pre-to-post-event reductions in the perceived fairness of gender relations. These findings suggest that the system-justifying effects observed in past work fail to emerge when specific, confronting experiences provide salient information that the system is not fair, does not work, and puts women in danger.

Women who reported sexual harassment also experienced reductions from their pre-event perception of the broader society as fair and just (general system justification). This effect was smaller than the reduction in gender-specific system justification, which aligns with the gendered nature of sexual harassment. The impact was, however, still evident for more general perceptions of society. Thus, discriminatory experiences in specific domains not only promote change within domain-specific beliefs, but also may generalize to broader beliefs about society that likely have important implications for social change. Women who perceive gender relations as unfair are more likely to oppose or seek to change the social system (see Jost et al., 2001). For example, Becker and Wright (2011) found that women exposed to hostile sexism were more likely to engage in collective action (e.g., participate in protest, sign a petition) because they perceived gender relations as less fair. More generalized perceptions that society is unfair may also reduce the degree to which women accept the status quo and increase their motivation to address inequalities. For example, Cichocka et al. (2018) found that lower and moderate (versus higher) support for the status quo is associated with participation in collective action (e.g., vote in an election, participate in protest). The current results, therefore, suggest that pivotal experiences may disrupt system justification processes by encouraging women to reject rather than accept the unfair (gendered) system and perhaps help combat the processes that maintain (gender) inequality.

Practice Implications

Although the results provide new evidence that gender-related experiences may raise awareness of the unfair nature of gender relations, this process likely only occurs if women are able to recognize or label experiences as sexual harassment. Given space constraints in the NZAVS, the BISLE uses a simple checklist probe to assess whether women experienced sexual harassment in the past year (yes or no; see Howard et al., 2022), which requires women to label, recall, and report an experience as sexual harassment (Chawla et al., 2021). However, women’s reports may depend on the type of experience, with unwanted sexual attention less likely to be recognized or reported as sexual harassment than more severe forms (Munson et al., 2000). Although meta-analytic evidence indicates that the effects of sexual harassment on well-being do not differ between single-item questions and more intensive behavioural checklists (see Chan et al., 2008), the BISLE measure is likely a conservative indicator of women’s sexual harassment experiences (Chawla et al., 2021). Relevant experiences not classified by participants as sexual harassment may not have the same effects and could be overlooked to protect well-being (consistent with system justification perspectives). Future studies that examine different types and interpretations of experiences may illuminate the salient experiential aspects that help women identify sexual harassment and, in turn, acceptance versus rejection of unfair gender relations.

Given that women who reported sexual harassment experienced reductions from their pre-event support for the unequal gender hierarchy, raising women’s awareness of the types of experiences that constitute sexual harassment may create meaningful social change toward gender equality. The current data were collected following the popularisation of #MeToo in 2017 and highlight the importance of such movements that increase understanding and challenge sexual harassment toward women. Recent research by Caputi et al. (2019) supports this claim, showing that, following the rise of #MeToo in 2017, web searches across the US increased not only for sexual harassment in general, but also for how to report and conduct preventative training for sexual harassment. This empirical evidence highlights the importance of raising awareness of sexual harassment to mobilize women to reject the unfair systems that support sexual harassment. Combined with the results in the current study, this evidence also emphasizes the importance of ensuring women have access to resources and support to cope with the declines in well-being that likely accompany experiences of sexual harassment.

Limitations and Future Research Directions

The current study advances the literatures on sexual harassment and system justification by examining pre- and post-event differences across women who reported sexual harassment in the past year and a matched control group with comparable outcome levels and demographics in the previous year. This approach allowed us to uniquely assess between-person (reported sexual harassment vs. matched control) and within-person (pre- vs. post-event) change. However, data constraints meant that we could not test the temporal ordering of variables, which requires more extensive longitudinal designs (see Bahamondes et al., 2021). It is likely that the links between sexual harassment and the outcomes we examined are bidirectional—well-being and system justification both shape and are shaped by reports of sexual harassment (see Napier et al., 2020). Indeed, our pre-matching analyses revealed that women who reported sexual harassment had lower pre-event levels of well-being and system justification than did all women who did not report sexual harassment. Crucially, however, we mitigated these selection effects by identifying a control group with comparable levels of pre-event well-being and system justification. Comparing post-event and pre-to-post-event changes to matched controls who had comparable baseline levels provides strong evidence that reports of sexual harassment are accompanied by declines in well-being and system justification.

The current study highlights the important socio-political implications of sexual harassment among women by showing that women who report sexual harassment are more likely to reject the unfair (gender) hierarchy. Applying system-justifying principles, we focused on two outcomes that often reflect attempts to protect (or reject) differential power relations between men and women (gender-specific system justification; see Jost & Kay, 2005) and general support for the unequal status quo (general system justification). Although the outcomes we examined correlate with collective action (see Osborne et al., 2019), we were unable to directly assess women’s support for, or participation in, gender-based or non-specific collective action. Engaging in collective action, such as tweeting about sexism, can promote well-being among women (see Foster, 2015). Thus, the drops in system justification shown in the current study may motivate behaviours that help to redress the decrements to well-being experienced following sexual harassment. We hope our novel results propel future research to assess how sexual harassment impacts not only women’s perceptions of gender relations and society, but also the flow-on decisions and behaviour that may ultimately produce change in (gender) inequality and improve women’s quality of life.

Although we assessed the impact of sexual harassment on women’s well-being and system justification separately, these two variables may interact to shape women’s lives. According to system justification theory, believing that society and gender relations are fair may serve a palliative function in which women are able to maintain their well-being in the face of adversity (see Bahamondes et al., 2019; Becker & Wright, 2011; Napier et al., 2020). Thus, reductions in women’s perceptions of gender relations and society as fair following sexual harassment may, at least in part, explain why these women also experience reductions in well-being. Although this hypothesis is theoretically plausible (see Bahamondes et al., 2019), the current results demonstrate that the palliative effects of system justification likely only go so far—highly salient and potentially traumatic experiences like sexual harassment are likely to reduce women’s system justification. We hope these results inspire future research to investigate the boundary conditions of the palliative function of system justification to advance understanding of when people are more likely to support or reject the status quo.

Conclusion

The current study utilized propensity score matching to examine pre-to-post-event changes in well-being and system justification across women who reported sexual harassment in the past year and a matched control group who were comparable in outcomes and demographics the year before the event but did not report sexual harassment. Strengthening prior evidence of the impact of sexual harassment on well-being, women who reported sexual harassment experienced significant declines from their pre-event levels of well-being. Notably, this pattern was a significant departure from the pre-post changes found amongst the matched controls. Advancing prior assessments and understanding of the impact of sexual harassment, women who reported sexual harassment also exhibited significant pre-to-post-event reductions in perceptions that gender relations, and broader society, are fair. These novel results suggest that pivotal gender-related experiences may encourage women to reject unfair gender relations and disrupt system justification processes that maintain (gender) inequality.