Introduction 

Sexual harassment (SH), as a form of gender violence, is an everyday occurrence worldwide. But in many countries, SH is overlooked or not even acknowledged as a form of violence. SH behaviors include non-verbal, verbal, and physical sexual coercive actions (Madan & Nalla, 2016; Neupane & Chesney-Lind, 2014). Common examples include unwanted sexual comments, lewd looks, and groping. However, the behaviors associated with SH can be as trivial as name-calling or as severe as rape (Nalla, 2020).

Prior research has investigated various facets of SH. The first is the space and location of its occurrence, which may be indoors or outdoors (the latter is commonly referred to as “street harassment”). Research also shows that it occurs in various contexts, such as in school or work (Barling et al., 2001; Fineran & Bennett, 1999; Richman et al., 1999), in the military (Fitzgerald et al., 1999), in public spaces (Ceccato & Paz, 2017; Korovich et al., 2021; Lichty & Campbell, 2012; Macmillan et al., 2000; Madan & Nalla, 2016; Neupane & Chesney-Lind, 2014; Rosenbaum et al., 2020; Shibata, 2020), and in the cyberspace (Barak, 2005; Barnes, 2001). A second area of research has focused on the impact of SH on the victims’ physical and mental health. SH victims experience an array of negative physical and mental consequences, including substance abuse, sleeplessness, eating disorders, anxiety, depression, and post-traumatic stress disorder (e.g., Dansky & Kilpatrick, 1997; Fileborn & O’Neill, 2021; Hill & Silva, 2005; O’Hare & O’Donohue, 1998; Richman et al., 1999; Shinsako et al., 2001; Street et al., 2007).

Although the public perceptions are more homogenous regarding more severe forms of gender violence such as rape and sex trafficking, there may be less consensus over their attitudes toward sexual harassment—a legally “lesser” offense. However, that, by itself, does not make SH a less severe violence. Therefore, understanding attitudes toward SH (specifically that against women, the primary targets of SH), as well as the public’s level of tolerance for SH, may help us better understand social attitudes toward gender equality. This is particularly of interest in some countries, such as China, where progress toward gender equality is hard to assess. On the one hand, many grassroots activists are calling for greater gender equality via social media campaigns; on the other, the Chinese government and government-backed organizations are suppressing these efforts (Han, 2018; King et al., 2013; B. Wang & Driscoll, 2019; Zeng, 2020). For instance, the Communist Youth League of China (Gong Qing Tuan, a branch of the Chinese Communist Party) called feminism “a poisonous tumor” publicly (see Yan, 2022). Furthermore, the media silenced the news about a trafficked and abused woman in Feng County, Jiangsu Province, despite the public uproar (e.g., Ho et al., 2022; Kuo & Li, 2022; USA Today, 2022).

The focus of the present research is on public tolerance for SH behaviors. Prior research has examined tolerance for SH in the context of previous sexual victimization (Reilly et al., 1992), gender roles (Russell & Trigg, 2004), organizational characteristics (Hulin et al., 1996), the impact of videogames (Dill et al., 2008), and within different professions (e.g., army, Rosen & Martin, 1998; athletes, M. Ahmed et al., 2018; lawyers, Zvi & Shechory-Bitton, 2022; medicine, Vargas et al., 2020). Research on perceptions of SH has also been extended outside North America and Europe, such as in India (Madan & Nalla, 2016; Nalla, 2020), Nepal (Neupane & Chesney-Lind, 2014; Rosenbaum et al., 2020), Bangladesh (Banarjee, 2020), South Korea (Jang & Lee, 2013; Y. Lee et al., 2013), Pakistan (B. Ahmed et al., 2021), and Singapore (S. Li & Lee-Wong, 2006).

Research on sexual harassment has also been conducted in China, but it is limited. Awareness of SH in China is still relatively new, given that the sexual revolution did not take place very long ago (Farrer, 2000). The research largely focuses on women’s SH experiences and the role of traditional Chinese culture on the risk and aftermath of sexual harassment victimization (S. Choi et al., 2016; Liu et al., 2014; Lui, 2016; Parish et al., 2006; Tang, 1994). The few that did study people’s attitudes toward sexual harassment found women to be less tolerant of SH because of their more egalitarian gender role beliefs (Tang et al., 1995a, 1995b). The current study is thus a much-needed investigation of the public’s tolerance of SH in today’s China.

This research contributes to the SH literature in several ways. First, relatively limited research on sexual harassment attitudes has been conducted outside the White-dominant and English-speaking Global North. Because of the cultural, historical, and contextual differences between China and the West, it is unclear whether the findings generated in prior research apply to the Chinese context. Second, this study explores differences, if any exist, in the tolerance for a range of SH behaviors between resident Chinese and Chinese citizens living abroad. More specifically, we draw from theoretical work on patriarchy and traditional gender role beliefs in shaping attitudes about sexual harassment. In addition, we use Sykes and Matza’s (1957) neutralization theory to test Chinese nationals’ tolerance for SH.

Definition of Sexual Harassment

Sexual harassment (SH) refers to a range of behaviors that are directed toward another person without that person’s consent. There are several different categorizations of sexual harassment. However, most are based on behavioral traits, such as whether the harassment takes the form of verbal comments, nonverbal cues, or physical actions (e.g., Gruber et al., 1996; Madan & Nalla, 2016; Nalla, 2020; Neupane & Chesney-Lind, 2014; Shibata, 2020; Timmerman & Bajema, 1999). In the current study, we grouped SH into three categories based on its purposes rather than its means of perpetration: unwanted sexual attention, gender harassment, and sexual coercion (Fitzgerald et al., 1988; Fitzgerald et al., 1995; Gelfand et al., 1995).

Unwanted sexual attention is the traditional form of “sexual harassment,” referring to behaviors for sexual attraction and gratification (Franke, 1997; Pina & Gannon, 2012; Schultz, 1998; Stockdale et al., 1995; Tangri & Hayes, 1997). This type of sexual harassment relies on verbal and nonverbal behaviors of a sexual nature, including sexually suggestive language and comments about someone’s appearances. As the name suggests, unwanted sexual attention is perceived by the victims as “unwelcome, unreciprocated, and offensive” (Page & Pina, 2015, p. 74). A potential explanation is that men tend to misperceive, or over-perceive, women’s friendliness as promiscuity or exhibiting romantic and sexual interest toward them (Abbey & Melby, 1986; Perilloux et al., 2012; Stockdale, 1993). Others note that unwanted sexual attention is shaped collaboratively by people’s understanding of “normative and ‘reasonable’ sexual expression” and “structural norms regarding gender and sex” (Fileborn, 2016, pp. 322–323).

Second, gender harassment is a provocation based on gender norms (Berdahl, 2007; Dall’Ara & Maass, 1999; Fiske & Stevens, 1993; Maass et al., 2003; Prentice & Carranza, 2002; Pryor & Fitzgerald, 2003) wherein males are to be masculine (e.g., strong and dominant) and women ought to be feminine (e.g., modest and submissive). Thus, if women do not exhibit specific “feminine” characteristics, they must be “put in their place” (Berdahl, 2007). In other words, gender harassment intentionally seeks to make the victim feel unwelcome, especially in traditionally male-dominated spaces (Page & Pina, 2015; Vera-Gray & Kelly, 2020). Similarly, men are more likely to harass when their masculinity is challenged or threatened and subsequently needs to be validated. Gender harassment against women is thus a public display and reinforcement of masculinity to both the perpetrator himself and other men (Berdahl et al., 2018; Maass et al., 2003; Schrock & Schwalbe, 2009).

Finally, sexual coercion is harassment that is rooted in sexual purposes and involves power differentials, where the perpetrator exercises social power over the victim to achieve sexual cooperation (Fitzgerald et al., 1995; Gelfand et al., 1995; McDonald, 2012; Page & Pina, 2015). Sexual coercion has a stronger focus on power (Bohner et al., 1993; MacKinnon, 1979; Schatzel-Murphy et al., 2009) and is thus different from unwanted sexual attention, which emphasizes sexual gratification. Furthermore, sexual coercion is often considered the more severe form of SH (O’Connell & Korabik, 2000). Although sexual coercion relies on power and sometimes involves physicality, it differs from sexual assault and rape (Testa & Derman, 1999). Sexual coercion is characterized as coercing victims into compliance in exchange for awards or to avoid negative consequences and is often done through extortions and threats (i.e., quid pro quo harassment; Fitzgerald et al., 1995). The focus, therefore, is on the coercive process rather than the sexual attempts.

Tolerance of Sexual Harassment

Gender Ideology and Tolerance

Opinions about gender shape people’s tolerance of SH; people endorsing more traditional and patriarchal gender roles will be more tolerant (e.g., Begany & Millburn, 2002; Foulis & McCabe, 1997; Herzog, 2007; Lonsway et al., 2008; Pryor et al., 1995; Reilly et al., 1992; Russell & Trigg, 2004). Specifically, Lonsway et al. (2008) find that common myths surrounding SH are strongly and positively associated with sexism, stereotypical sex roles, and hostility toward women but negatively with support for feminism. Although both men and women share sexist myths, the more one believes in these traditions, the more tolerance they have for sexual harassment (Herrera, et al., 2017; Lonsway et al., 2008) and the more likely they are to engage in victim blaming (Lila et al., 2013). Research that compared the effects of sex and gender beliefs on SH attitudes shows that although some differences exist between men and women’s tolerance for sexual harassment behaviors, endorsement of sexism and stereotypical gender roles is a much stronger determinant of tolerance than simply biological sex (Russell & Oswald, 2016; Russell & Trigg, 2004).

Techniques of Neutralization and Tolerance

First suggested by Sykes and Matza (1957), the techniques of neutralization suggest five methods commonly used by offenders to justify their deviance (Sykes & Matza, 1957): denial of responsibility (i.e., that the act is a result of circumstantial factors or force); denial of injury (i.e., that the act itself does little or no harm); denial of victim (i.e., that there is no direct victim of the act or the victim deserves the harm); condemnation of condemners (i.e., the offender shifts the focus onto the rejectors of their deviance); and appeal to higher loyalties (i.e., that the act is done to please or fulfill an organizational, holistic, or other mission).

Although Sykes and Matza (1957) proposed the techniques of neutralization to explain juvenile delinquency, the theory is often relied upon as a theoretical explanation for sexual and gender violence (Boyle & Walker, 2016; Cavanagh et al., 2001; DeYoung, 1988; Dutton, 1986; Henderson & Hewstone, 1984; Hinze, 2004; Hlavka, 2014; Hubner, 2008; Rolfe & Schroeder, 2020; Scully & Marolla, 1984; Shakeshaft et al., 1997; Spraitz & Bowen, 2016; Stubbs-Richardson et al., 2018). The two techniques of denial of injury and denial of responsibility are particularly prevalent in the research (e.g., Boyle & Walker, 2016; Hinze, 2004; Hlavka, 2014; Hubner, 2008; Rolfe & Schroeder, 2020; Shakeshaft et al., 1997). For example, research on intimate partner violence often finds that male abusers do not accept their responsibility. Instead, they utilize neutralization techniques to attribute responsibility for their violence to their female victims/spouses (Cavanagh et al., 2001; Dutton, 1986; Henderson & Hewstone, 1984) or the influence of drugs and alcohol (Lavin-Loucks & Levan, 2015; Lehrner & Allen, 2008; Scully & Marolla, 1984). In contrast, however, fewer studies have examined SH within the neutralization framework, and even fewer studies have attempted to apply the theory to crime and delinquency in Chinese societies (He, 2015; J. Li & Wu, 2012; Zhang et al., 2017).

Individual Factors and Tolerance

Prior studies generally find that men are more tolerant of SH than women. Specifically, men perceive sexual harassment to be less serious in general and are more likely to overlook sexual harassment behaviors (e.g., Beauvais, 1986; Berdahl, 2007; DeSouza & Fansler, 2003; Dill et al., 2008; Gutek, 1985; Lonsway et al., 2008; Mazer & Percival, 1989; O’Leary-Kelly et al., 2009; Shechory-Bitton & Shaul, 2013 Tata, 1993). For example, Shechory-Bitton and Shaul (2013) find that women have a much broader and more gender-egalitarian definition of SH and they are more likely to recognize behavior as SH regardless of the victim and harasser’s sex.

Although earlier research in the Western contexts finds younger adults (such as college students) to be more tolerant of SH than older respondents (e.g., Baker et al., 1990; Foulis & McCabe, 1997; Lott et al., 1982), contemporary literature largely zooms in on gender divisions within young adults’ attitudes and seldom compares attitudinal differences across age groups. One of the few recent studies reported that older adults (especially older men) tend to feel more positively about political figures accused of sexual harassment and assault (Hansen & Dolan, 2020). The effects of age and education are even less known in the Chinese context (e.g., Jiang et al., 2009, 2018). We believe that college students today are more conscious of social justice issues (Ferguson & Ireland, 2006; Hanel & Vione, 2016; Maggard et al., 2012; Payne et al., 2006; Peterson & Merunka, 2014; Williams & Nofziger, 2003) and, therefore, less tolerant of sexual harassment—especially in the case of China, where college students are typically the most progressive and at the frontline of social change (Q. Wang, 2018), as evidenced by recent events surrounding feminism and other social justice issues (e.g., prominent figures include Yue Xin in the 2018 JASIC labor dispute).

Chineseness and Tolerance

There is a significant lack of social recognition of SH in China, and all parties involved—victims, perpetrators, and the public—largely overlook its seriousness. The reason for this is the traditional Chinese notion of patriarchy and stringent gender roles to which victims, perpetrators, and bystanders are accustomed (Chan, 2009; Luo, 1996; Ruan, 1991; Shi & Zheng, 2020; Tang et al., 1995a, b, 2000). For example, it is estimated that only 10% of rapes are reported in China (Luo, 2000), and many consider the notion of “face” to blame (Chan, 2009; Hu, 1944; M. Lee & Law, 2001; Xue et al., 2019). People also tend to think the victim is at least partially to blame (J. Lee et al., 2005; Xue et al., 2019). Moreover, female chastity is highly valued in Chinese societies historically, and a woman is only supposed to engage in sexual behaviors with her legal husband. Any violation of this virtue, even in the case of rape and sexual assault, is regarded as a loss of face (for the family) and therefore worthy of intense and prolonged stigmatization (L. Li, 1999). In this sense, coming forward with any sexual victimization is considered a violation of Chineseness and a threat to the empowerment of China, also seen in the ongoing controversy over China’s #MeToo movement, where activists are labeled “spies” and “traitors” for speaking out against sexual harassment (Lin & Yang, 2019). Therefore, it is reasonable to predict that people outside mainland China are more vocal and less tolerant of SH, because they are exposed to different cultures and given the freedom to speak up.

Given the limited scholarly attention paid to the public perceptions of SH in the Global South, this research seeks to fill the gap by looking exclusively at Chinese nationals’ attitudes toward SH of different kinds by addressing the question: How do Chinese people think of sexual harassment? Drawing from the literature review, we hypothesize the following:

  • Hypothesis1: Endorsement of traditional gender roles increases people’s tolerance of SH.

  • Hypothesis2: Endorsement of patriarchal gender expectations increases people’s tolerance of SH.

  • Hypothesis3: People who deny injury in SH cases are also more tolerant of SH.

  • Hypothesis4: People who deny responsibility in SH cases are also more tolerant of SH.

  • Hypothesis5: Women have a lower tolerance of SH than men.

  • Hypothesis6: Students have a lower tolerance of SH than non-students.

  • Hypothesis7: Chinese nationals living abroad have a lower tolerance of SH than those living in China.

The Current Study

The Chinese Context

The Chinese Law of Protecting Women’s Rights and its regional modifications define sexual harassment as a verbal, graphic, or physical act against a woman’s will. Under Article 6 (Section 40) of the Chinese Law of Protecting Women’s Rights, sexually harassed women can file complaints to relevant organizations and government branches. Despite these laws, however, SH perpetrators in China rarely receive formal sanctions. And as a comparison, the minimum sentence for rape is three to ten years in prison and can even be as severe as the death penalty (Criminal Law of the People’s Republic of China Article II, Section 236).

Socially, SH has been labeled a “hush-hush” topic in Chinese societies, and research on the phenomenon did not begin until the late twentieth century (Luo, 1996; Tang et al., 1995a, b, 1996). Even though women are more alert to, and can more easily recognize SH (Tang et al., 1996), victims in China often avoid seeking help and allow the perpetrators to escape consequences because of guilt and shame (Chan et al., 1999; Dussich, 2001; Tang, 1994; Tang et al., 1995a, 1995b). Other researchers find that victims are inclined to shift the responsibility to the community or criminal justice institutions to act against SH (Tang et al., 1996); in other words, while they call for systematic changes and community efforts, they do not want to speak up for themselves. Overall, sexual harassment remains a taboo topic in contemporary China.

Data

The data used in the current study is derived from a project comparing attitudes toward SH across different cultural groups. In the spring of 2020, we created an online survey targeting Chinese and US nationals composed of three parts: (1) attitudes about gender, (2) experiences with SH training, and (3) perceptions of sexual harassment. The questionnaire was created in English but later translated into Simplified Chinese by the lead author of this article and back-translated by a group of five bilingual individuals outside social sciences. This ensured that survey items were free of academic jargon and made sense in the Chinese language and that people using different languages were answering the same questions. The survey link was created on Qualtrics and distributed to both Chinese and US nationals through convenience sampling (through social media platforms, direct emails and listservs, university resources, and professional networks). A total of 2943 individuals opened the survey, but approximately one-tenth of them dropped out immediately. Most respondents completed the entire survey despite skipping some questions; the response rate is between 69.5 and 89.7% for all variables included in this study. We also include only Chinese nationals in the current analyses (N = 2094). The descriptive statistics for the variables used in the study are presented in Table 1. A vast majority of the respondents in our study are women (88.7%), students (82.9%), and residing in China (88.2%).

Table 1 Descriptive statistics (N = 2094)
Full size table

Measures

Dependent Variables

The questionnaire borrows from the Sexual Experience Questionnaire ([SEQ]; Fitzgerald et al., 1988; Gelfand et al., 1995) to measure respondents’ attitudes toward SH. We identify a total of 12 items that are commonly present when inquiring about attitudes about SH and ask respondents to answer the question, “Do you think [item] is acceptable?” on a five-point Likert scale from “strongly unacceptable” (1) to “strongly acceptable” (5).

The 12 items are found in different constructs in the SEQ; therefore, we conducted a factor analysis to determine the different latent constructs. Two items (“tell suggestive stories” and “invasion of privacy”) are dropped due to their low factor loadings (< 0.55). Results confirm that three categories of harassment exist in the data: unwanted sexual attention (five items), gender harassment (three items), and sexual coercion (two items). Unwanted sexual attention (UST) includes behaviors such as asking for a hookup (casual sexual encounter), deliberate touching, attempts to fondle, bribe for sexual cooperation, and reward before sexual cooperation (α = 0.781). Gender harassment (GH) includes behaviors such as differential treatment because of gender, making sexist remarks, and displaying sexist materials (α = 0.668). Lastly, sexual coercion (SC) includes making threats for lack of sexual cooperation and negative consequences for refusing sex (α = 0.694).Footnote 1 For each of these three dependent variables, cumulative scales are created, and their average scores are calculated and used in the statistical analyses.

Independent Variables

The current study includes three sets of independent variables comprising control, gender beliefs, and use of neutralization techniques. First, we consider the effect of sex (0 = man, 1 = woman), location of residence (0 = in China, 1 = abroad), and whether the respondent is a college student (0 = non-student, 1 = student).

Second, using statements from the Attitudes toward Women Scale ([AWS]; Spence et al., 1973), we measure respondents’ attitudes related to gender. Factor analysis confirms that two separate constructs are measured in the 11 items, which we label “traditional gender roles” and “patriarchal gender expectations” (three items with factor loadings below 0.55 were dropped from the analysis). The traditional gender roles (TGR) include four items about women’s “place” and agency, such as their ability to divorce, domestic labor distribution, and the role of sex in the job market (α = 0.629). Patriarchal gender expectations (PGE) include four statements: “sons should be given more encouragement to go to college than daughters,” “fathers should have more authority in bringing up children,” “girls should not engage in sexual activities before marriage,” and “women should worry more about becoming good wives and mothers.” All four items relate to societal expectations of men and women (α = 0.669). For each of the two variables, cumulative scales are created, and average scores are calculated and used in the analyses. The sample has a relatively low subscription of patriarchal beliefs in both their endorsement of traditional gender roles (mean = 1.31, S.D. = 0.43) and their patriarchal gender expectations (mean = 1.58, S.D. = 0.61; Table 1).

Last, we include two neutralization techniques that are common in sex offending and gender violence research, namely denial of injury and denial of responsibility (e.g., Boyle & Walker, 2016; Hinze, 2004; Hlavka, 2014; Hubner, 2008; Rolfe & Schroeder, 2020; Shakeshaft et al., 1997). Denial of injury is the argument over “whether or not anyone has clearly been hurt” (Sykes & Matza, 1957), and the goal is often to lessen the seriousness and the consequences of the deviant behavior (Harris & Dumas, 2009; Maruna & Copes, 2005). The questionnaire asks respondents to rank four types of sexual harassment (i.e., verbal, nonverbal, physical, and serious physical; Madan & Nalla, 2016) based on their perceived seriousness. If a respondent ranked “serious physical harassment” (e.g., rubbing breasts, poking with a penis) as anything but the most serious, we considered it to be minimizing harm and downplaying the significance of the behavior, and therefore denial of injury (DOI; 0 = no, 1 = yes). About one-third of the sample adopt this technique (Table 1).

In denial of responsibility, the perpetrator asserts that their actions are due to external forces beyond their control (Agnew & Peters, 1986; Sykes & Matza, 1957). One everyday use of this technique is attributing blame to substance use, such as alcohol and drugs (Boyle & Walker, 2016; Lavin-Loucks & Levan, 2015; Lehrner & Allen, 2008; Scully & Marolla, 1984; Spraitz & Bowen, 2016). The questionnaire asks respondents to rate whether one is more likely to sexually harass others when they are “sober” or “under the influence.” If they consider people “under the influence” to be more likely to offend, they are considered endorsing denial of responsibility (DOR; 0 = no, 1 = yes). Close to two-thirds of the sample consider “being under the influence” a liable reason for sexually harassing others (Table 1).

Results

Multivariate Regression Models

We rely on ordinary least square (OLS) models in the current study to estimate predictors of tolerance for SH behaviors among Chinese nationals. Collinearity statistics in all models suggest that multicollinearity assumptions were not violated—the variance inflation factors (VIFs) were below 1.5.

Predictors of Tolerance in Sample

Table 2 displays the regression models on predictors of SH tolerance in the full sample (N = 2094). The independent variables in this model explain a 9.4% variance in tolerance for unwanted sexual attention, 23.9% for gender harassment, and 13.1% for sexual coercion. Compared to men, women’s tolerance of SH is consistently lower for all three subtypes of SH. The location appeared to influence tolerance levels for at least one of the three forms of SH. Compared to respondents who live in China, those abroad have a greater tolerance for unwanted sexual attention (β = 0.07, p < 0.001).

Table 2 Predictors of tolerance in sample (N = 2094)
Full size table

Findings from Table 2 suggest a strong relationship between respondents’ gender ideology and tolerance for SH. Those who believe in traditional gender roles are more tolerant of all subtypes of SH. A similar relationship was found between respondents’ perceptions of patriarchal gender expectations and tolerance for gender harassment (β = 0.27, p < 0.001) and sexual coercion (β = 0.22, p < 0.001).

The findings from this study show that the impact of neutralization techniques is minimal in explaining Chinese citizens’ tolerance for SH. Respondents who minimize the harm in serious SH incidents are also more tolerant of unwanted sexual attention (β = 0.05, p < 0.05). However, no significant finding was established for the remaining two forms of SH. Denial of responsibility did not predict the respondents’ tolerance for SH behaviors at all.

Additional Analysis: Predictors of Tolerance Between Students and Non-students

Research in the Western context suggests that college students tend to differ from the general population in their attitudes toward social justice issues (e.g., Ferguson & Ireland, 2006; Hanel & Vione, 2016; Maggard et al., 2012). Although such differences are not well explored in research on Chinese populations (e.g., Jiang et al., 2009, 2018), we cannot exclude the risk of misinterpreting the results and making misinformed inferences with the combined sample.

We conducted multivariate analyses for students (N = 1728) and non-students (N = 357) separately (Table 3). The adjusted R2 for non-students and students were 14.4% and 8.8% for unwanted sexual attention (model A), 31.3% and 22.9% for gender harassment (model B), and 9.8% and 13.8% for sexual coercion (model C), respectively.

Table 3 Predictors of tolerance in students (N = 1728) and non-students (N = 357)
Full size table

Sex is the strongest predictor of tolerance for both non-student and student groups across all three forms of sexual harassment. However, compared to non-students, students living abroad have more tolerance for unwanted sexual attention (β = 0.07, p < 0.01). There were no statistically significant differences between the two groups found for the other two categories of SH.

Consistent with expectations, endorsement of traditional gender roles predicts people’s tolerance for all categories of SH behaviors in both groups of respondents. Belief in patriarchal gender expectations is related to non-students and students’ tolerance of gender harassment (β = 0.34, p < 0.001 vs. β = 0.24, p < 0.001); however, it only predicts students’ tolerance of sexual coercion (β = 0.20, p < 0.001). And surprisingly, the less expectation a non-student has for women’s potential and contribution to society, the less tolerant they are of men’s unwanted sexual attention toward women (β = − 0.11, p < 0.1).

Next, neutralization techniques are not found to be significant factors in predicting either students’ or non-students’ tolerance of sexual coercion. The use of denial of injury likely raises students’ tolerance of unwanted sexual attention (β = 0.07, p < 0.01), but it reduces non-students’ tolerance of gender harassment (β = − 0.14, p < 0.01). Denial of responsibility, however, was not found to be related to any of the dependent variables. Overall, techniques of neutralization had limited explanatory power in clarifying respondents’ tolerance for various forms of SH behaviors.

Discussions

The goal of our research is to examine tolerance for three forms of SH behaviors among Chinese nationals living in China and abroad. We also compare the attitudes between students and non-student Chinese nationals. We drew our research hypotheses from two key theoretical frameworks, namely patriarchy and gender roles, and Sykes and Matza’s neutralization theory.

One of the most salient findings in our study is that endorsement of traditional gender roles positively predicts people’s tolerance of all types of SH, which supports prior research findings in Western societies (e.g., Berdahl, 2007; Dall’Ara & Maass, 1999; Lonsway et al., 2008; Maass et al., 2003; Russell & Trigg, 2004) and lends support for Hypothesis 1. In contrast to beliefs in traditional gender roles, the effect of subscription to patriarchal gender expectations varied considerably, lending partial support for Hypothesis 2. In summary, the results somewhat support that gender role ideologies shape people’s views on SH in China (e.g., Tang et al., 1995a, b). Although prior research in China has found power differentials to be only moderately related to attitudes about SH (Liu et al., 2014; Parish et al., 2006), with the rapid social transformations and increasing awareness of power and control in China, the results we report in the current paper should be expected.

Regarding the use of neutralization techniques, the effects of denial of injury (Hypothesis 3) and denial of responsibility (Hypothesis 4) all vary significantly depending on the type of SH as well as between students and non-students. The findings here are rather opposite to previous research suggesting that people neutralize sexual harassment as “a way of life” or a part of everyday culture and thus are more tolerant of SH (Hinze, 2004; Hlavka, 2014; Hubner, 2008; Rolfe & Schroeder, 2020; Shakeshaft et al., 1997).

Perhaps the strongest finding was support for Hypothesis 5, which relates to sex and tolerance for SH—that is, women are less tolerant of all three forms of SH. Consistent with previous studies in the Western contexts, we find Chinese women to be much less tolerant of all SH behaviors than men, regardless of whether they are current students (e.g., Beauvais, 1986; DeSouza & Fansler, 2003; Gutek, 1985; Shechory-Bitton & Shaul, 2013). This observation also corroborates contemporary events in China, where women are much more vocal and determined about feminism and equal rights, and men are comparatively silent or adamant about traditional gender values (as we have discussed in the introduction).

We find mixed support for Hypothesis 6, that students and non-students differ considerably—for example, location of residence is a predictor only among the students and the effect of subscription to patriarchal gender expectations goes in opposite directions for students and non-students—we find some evidence that college students may differ from non-students in their attitudes toward social justice issues (Ferguson & Ireland, 2006; Jiang et al., 2009, 2018; Maggard et al., 2012; Payne et al., 2006; Williams & Nofziger, 2003). However, because the two subgroups remain largely similar in what predicts their tolerance for sexual harassment overall, we would like to suggest that, with adequate precautions, generalizing findings from college samples to make inferences about the larger populations might be a viable solution for research on Chinese societies, given the large population size.

Last, we reject Hypothesis 7, which proposes that Chinese nationals residing outside mainland China are less tolerant for SH. We find that Chinese nationals residing abroad tend to have higher tolerance of sexual harassment behaviors (specifically unwanted sexual attention) than those in China, which is against the assumption that acculturation to the “liberal” West lowers the individual’s acceptance of sexual harassment cultivated in their conservative and patriarchal home countries (e.g.,Bejarano et al., 2011; Cortina & Wasti, 2005; Kennedy & Gorzalka, 2002; Nalla, 2020; Tang et al., 1995a, b). However, similar findings are shared by a study of Indian migrants (Rajan et al., 2020), who hold much more conservative gender attitudes than non-migrants irrespective of where they return from. Many factors can potentially explain this finding, such as that their conservative views are resistant to change even when one migrates to more liberal environments (Inglehart, 1997). Especially at present (while drafting this manuscript, the Roe v. Wade decision was overturned in the USA, thereby devolving the question of abortion’s legality to the individual states), it is debatable how liberal the West remains and what influence the Western culture has on Chinese migrants. Another potential explanation is that those migrating abroad remain conservative to the traditional values as a form of cultural loyalty to the homeland and fortification of their cultural identity, even though things are no longer the same at home (e.g., Portes & Rumbaut, 2001; Verkuyten & Yildiz, 2010).

Conclusion

In conclusion, the findings from our study suggested sex and gender ideologies are strong predictors of Chinese nationals’ tolerance of sexual harassment behaviors. In contrast, neutralization theory and location of residence were found to explain tolerance for only unwanted sexual attention, but not other forms of SH behaviors.

The findings may bear some practical implications. For example, because gender roles significantly influence respondents’ attitudes toward sexual harassment, it is worth some consideration for educational institutions—K-12 schools, universities, non-government organizations, and educational branches of the government—to incorporate gender equality into their curriculums. It is also necessary to give women a voice, if not more, in current social and political affairs. Because women are more progressive, as reported in our study, they are more passionate and competent to find solutions to social justice issues, compared to men who may be slow and reluctant to change.

Although the findings are interesting and constitute a novel contribution to the existing literature, we must call attention to a few limitations of the study. For one, we include only a limited number of independent variables derived from established theories, and the analyses could benefit from a stronger theoretical framework. Second, relying on a convenience sample may mean that we face a greater likelihood of bias in the analyses and interpretation of results. Thus, results from the current study should not be generalized to the entire Chinese population. Our sample also relies on a cross-sectional design and is primarily female, which likely magnifies this issue. Future research should attempt probability sampling and draw inferences based on nationally representative data through stratified sampling techniques, which improves the generalizability of the statistical results reported. Last but not least, some of the statistical results are contrary to our assumptions. For instance, we find those residing abroad to be more tolerant of harassment behaviors. Qualitative inquiries could be immensely helpful in making sense of the findings.