Harassment in Brazilian universities: how big is this problem? The Federal University of Rio Grande do Sul (UFRGS) as a case study

Abstract

Harassment is a common problem in many institutions, including academic ones. It creates invisible barriers for attracting and maintaining women and other minorities at academia. In this work, we quantify and qualify this problem in one of the biggest universities of Brazil, the Federal University of Rio Grande do Sul (UFRGS). We applied a spontaneously-responded survey to 25% of its professors, 20% of its staff and 12.5% of its students. Our results present that sexual and moral harassment are very frequent in UFRGS and are not distributed homogeneously among all groups: women, black+parda, and non-heterosexual people are the most common victims. Moreover, the staff has higher percentage of moral harassment victims. We also unvealed an important problem of under-notification, where only around 10% of these cases are reported. We show that victims are afraid of reporting or do not believe in the institutional channels of reporting proposed by the university. We then discuss which type of events are perceived as harassment and the frequency at which they happen with different genders. We finalize the paper with a discussion of our results and recommendations to improve this scenario.

Key words
diversity in science; harassment in academia; women in science; sexual harassment; moral harassment

Introduction

The Brazilian public university, with a primary purpose of producing knowledge and training professionals capable of providing solutions to social problems and scientific challenges, is also a workspace which is not exempt from violence. In the study presented here, we seek to understand the perception of civil servants (professors and technical-administrative staff) and students regarding attitudes that lead to sexual and moral harassment. We also aim to explore the ways in which harassment is identified in the various spaces of the university, and which segments of the staff are the most affected.

Among the multiple existing definitions, sexual harassment may be defined as verbal and physical behaviors of a sexual nature associated to hostile and degrading attitudes (Martin-Storey et al. 2018). It can also be characterized as an abusive conduct toward obtaining sexual favors, humiliating, and disqualifying a person based on their sex, sexual orientation, or gender identity/expression. As Maria Ester Freitas recalls (Freitas 2001), sexual harassment is not flirtation, it is blackmail. Moral harassment, although often confused and linked to sexual harassment (mainly when it is motivated by gender), is defined as violent behaviors and attitudes that are based on humiliation, aggression, psychological terror and abuse of power, interfering in the health of professors, staff members, and students in their professional and training trajectories (Andrade 2018). It is important to remember that both types of harassment damage the dignity and health of the victim, in addition to jeopardizing their job/position and degrading their work environment (Hirigoyen 1998).

The studies cited above, among others, demonstrate a relationship between both forms of harassment and gender, race/color, origin, disability, sexuality, and hierarchical position in the institution. Moral harassment, also known as mobbing, is a concept often used in Latin America and Europe. It is a broader concept than sexual harassment, once it encompasses different forms of harassment associated with gender, sexuality, race, age, class, geographic origin, religion, etc. According to Freitas, the first important study about moral harassment was conducted by the swedish rechearcher Heinz Leymann (Leymann 1990). Other studies conducted with university samples, both qualitative and quantitative such as Caran et al (Caran et al. 2010), Guimarães, Cançado and Carvalho (Guimarães et al. 2016), and Cogenli and Barli (Cogenli & Barli 2013) have shown that victims of harassment have been predominantly women and sexual minorities, with race/color as an aggravating factor.

It is also noteworthy that sexual minorities (LGBT - Lesbian, Gay, Bisexual and Travesti ¹/Transsexual population) have been identified in the literature as more vulnerable to sexual harassment in University settings (Martin-Storey et al. 2018). In addition, especially in the Brazilian society, the racial dimension must be taken into account in the analysis, due to structural and systematic racism (Silveira et al. 2014).

In order for the possibility to fight harassment in University environments, in addition to its recognition, it is imperative to build institutional tools as training; improving university environment regarding gender, sexuality, and race; as well as creating effective reporting channels and accountability for harassers (Clancy et al. 2020). In this direction, and with the approval of the rectory, this study sought to draw a picture of moral and sexual harassment in a large federal university in southern Brazil. To this end, inspired by the Harvard University pioneer study (Kenig & Ryan 1986), an instrument was applied consisting of questions regarding socio-demographic and institutional information, and the “Harvard Sexual Harassment Survey“ (Verba et al. 1983). The instrument was sent to all segments of the university: professors (this category includes lecturers with short-term contract and assistant, associate, and full professors), staff (which includes technical and administrative professionals), and students (undergraduate and graduate). To participate in the research, the participants received an online access link available from May to June 2019 through email, with the Informed Consent Form authorizing their voluntary participation. 739 or 25.4% of the professors, 521 or 20.2% of the administrative staff and 4791 or 12.4% of the students (graduate and undergraduate) accepted to participate in the survey. The research was approved by the Research Ethics Committee of the Psychology Institute of the Federal University of Rio Grande do Sul (UFRGS) and integrates the actions of the HeForShe UFRGS² and Meninas na Ciência UFRGS projects.

Research questions

The goal of this study is to quantify the harassment problem at UFRGS and qualify some of its characteristics. In particular, the main questions guiding this study are the following: i) What is the percentage of victims of harassment in the academic community? ii) What is the profile of the victims and of the harassers? iii) What is the percentage of people who report this problem? Why do people not report? iv) What is perceived as moral and sexual harassment? Answers to these questions may help us to build, on the basis of quantitative data, policies for a more diverse and inclusive academic environment.

Profile of the respondents

Our sample size calculation was performed using Epi Info™ StatCalc version 7.2.1.0, “Sample Size and Power for Population Survey or Descriptive Studies” tool. Using the expected frequencies of harassment episodes in the "Harvard Sexual Harassment Survey", considering a margin of error of 5% and a confidence interval of 99%, the minimum sample sizes were calculated based on the actual number of teachers (2956), technical and administrative staff (2579), and students (44085) at the moment of data collection. It was established that at least 526 teachers, 512 administrative staff, and 628 students would be needed to present reliable measures of central tendency.

As we can see from the data in the Table I, the minimum sample sizes were met. This table also summarizes the profile of people who answered the questionnaire in terms of various characteristics: their sex assigned at birth (civil registration) and how they identify themselves now, their race/ethnicity, sexual orientation, lack of disability, and if the person is religious. We note that when it is just written "women" it refers to "cisgender women", and "men" refers to "cisgender men" throughout the paper.

Table I
The first line shows the total number of questionnaire respondents. The second one exhibits the percentage of the total number of teachers, staff, and students at UFRGS. The other lines present the profile of the respondents in the different characteristics asked in the questionnaire.

Total and Profile of the Respondents
		Professors	Staff	Students
Total	Number	739	521	4791
	Percentage	25,4%	20,2%	12,4%
Sex assigned at birth	Woman	55%	63,1%	63%
	Man	45%	36,9%	37%
Self-Identification	Woman	54,5%	63%	62%
	Man	45,4%	36,3%	36,3%
	Travesti	0,14%	0%	0,13%
	Trans Woman	0%	0,19%	0,17%
	Trans Man	0%	0%	0,23%
	Non-binary	0%	0,58%	0,75%
	Others	0%	0%	0,46%
Race Ethnicity	White	94,3%	85,3%	80,6%
	Black	1%	5,5%	7,2%
	Parda	3,8%	8,3%	11,4%
	Asian	0,85%	0,4%	0,52%
	Indigenous	0%	0,61%	0,3%
Sexual orientation	Heterosexual	90,5%	91,2%	74,1 %
	Homosexual	6,5%	3,8%	7,7%
	Bisexual	3%	3,8%	16,1%
	Asexual	0%	0,77%	0,71%
	Other	0%	0,38%	1,4%
Other characteristics	no Disabilities	96,8%	94,4%	98,6%
	no Religion	52,9%	44,7%	58,7%
	Average age	47 y.o.	42 y.o.	27 y.o.

We adopted the same race and ethnicity categories as IBGE (institute responsible for the population census in Brazil): branca (white), preta (black), parda, amarela (yellow, translated as Asian), and indígena (indigenous). There is no English translation of the term parda, which designates people with lighter skin tones when compared with the direct African descendants. They usually self-declare as black IBGE Educa (2020), Anteneodo et al. (2020).

Some profile characteristics of the respondents are noteworthy. Table I shows that, among them, 55% of the professors and approximately 63% of the staff and students are women, which correspond to a higher percentage of women in each category. Considering the semester data when the questionnaire was applied, the percentage of women in the total amount of people in each category is as follows: 46.5% of 2956 professors, 46.5% of 2579 staff members, and 51% of 38505 students.

In terms of race/ethnicity, we observed a large underrepresentation of black+parda people among professors. According to IBGE, the percentage of self-declared black+parda people in Brazil is around 54%. In the Rio Grande do Sul state, where UFRGS is located, this percentage is around 18%. Among respondents, there were less than 5% of black+parda professors. This percentage increases to 14% among the technical administrative staff and reaches 18.6% among students, a value more similar to the state average of this category. It is important to monitor these data over time to see if there is an increase in black+parda people between professors and technical administrative staff.

There is a significantly higher percentage of bisexuals among students compared to the other categories. Only 3% of professors and 3.8% of staff declared themselves as bisexuals, while this percentage rises to 16% among students. This change in sexual behavior between generations has been previously observed (Twenge et al. 2016, Athernon et al. 2016).

Another characteristic observed in our analysis of the university which is not representative of the Brazilian society is the percentage of people who have no religion. Among the general Brazilian population, approximately 14% declare themselves without religion (Inst. Humanitas 2020), while amid respondents there are around 50% in all categories who do so.

Results

In this section we first discuss the profile of moral and sexual harassment victims. Next, we present that very few people notify harassment and discuss some possible ideas to explain the lack of reporting. Afterwards, we explore the profile of the population who commit harassment, and we end this section presenting which types of episodes are perceived as sexual and moral harassment among the respondents, and which are more frequent.

To verify the statistical significance of our results, we measured the p value, ${\chi }^2$ and Cramer’s V. We consider p $<$ 0.05 the limit below which the differences between the frequencies in comparing groups are statistically relevant.

Part of these results can be read in a document written in Portuguese to present them to the UFRGS community. Along with the document, the three questionnaires prepared and applied can be obtained in Rosa et al. (2020).

Profile of the Victims

Table II presents the total number of responses among professors, technical administrative staff, and students who declared having suffered moral harassment. It also exhibits the percentage of respondents in each category who suffered harassment. One can easily observe that moral harassment is very common, reaching around 40% of professors and students and more than half of the technical administrative staff respondents.

Table II
First line shows the total number of people who suffered moral harassment among respondents to the questionnaire and the second line shows the percentage in terms of the total number of professors, technical and administrative staff and students. From the third line, the percentages of people harassed are discriminated in terms of different categories, whose identification is shown in the first column. For each of group (professors, staff and students) there are two columns: the column identified as ”R”, which corresponds to the percentage of respondents who suffered harassment normalized by the universe of respondents in each category and the column ”H”, that is the percentage of harassed people in each category. The last three columns are ${\chi }^2$ tests, p-values, and Cramer’s V for effect size.

Moral Harassment
	Professors		Staff		Students		${\chi }^2$	p	Cramer’s V
Total number	315		270		1851
Percentage	42,6%		51,8%		38,6%		35,94	$<$0.001	0.08
	R	H	R	H	R	H
Woman	51,9%	67%	56,5%	68,9%	44,7%	72,9%	22.20	$<$ 0.001	0.08
Man	31%	33%	43,8%	31,1%	28,3%	27,1%	19.86	$<$ 0.001	0.09
Woman	51%	65,9%	56,8%	68,9%	45,5%	71,8%	21.50	$<$ 0.001	0.08
Man	31,6%	33,7%	42,3%	29,6%	28%	26,3%	17.13	$<$ 0.001	0.09
Travesti	100%	0,32%	0%	0%	32%	0,11%	1.56	0.21	0.47
Trans Woman	0%	0%	100%	0,37%	50%	0,22%	0.9	0.34	0.32
Trans Man	0%	0%	0%	0%	36,7%	0,22%	-	-	-
Non-binary	0%	0%	100%	1,1%	44,3%	0,86%	3.42	0.64	0.29
Others	0%	0%	0%	0%	18,4%	0,22%	3.76	0.52	0.40
While	42,5%	94,8%	52,1%	84,3%	38,1%	79,9%	34.76	$<$ 0.001	0.08
Black	56,4%	1,3%	66,7%	6,9%	41,3%	7,8%	1.61	0.03	0.14
Parda	40,3%	3,6%	48,8%	7,7%	47,9%	11,5%	7.70	0.45	0.45
Asian	16,6%	0,33%	0%	0%	41,7%	0,6%	2.45	0.29	0.29
Indigenous	0%	0%	100%	1,2%	42,9%	0,3%	3.24	0.07	0.07
Heterosexual	41,2%	87,5%	50,9%	89,6%	35%	67,1%	49.87	$<$ 0.001	0.10
Homosexual	50,3%	7,7%	55%	4,1%	50,7%	10,1%	0.16	0.92	0.02
Bisexual	68,6%	4,8%	70%	5,2%	48,3%	20,2%	6.90	0.03	0.09
Asexual	0%	0%	50%	0,7%	47,1%	0,9%	0.01	0.91	0.02
Others	0%	0%	50%	0,4%	52,3%	1,8%	2.64	0.27	0.19

For each of these characteristics and each category we present two columns with different information: one that corresponds to the percentage of respondents – indicated in the table by "R" – in the given classification, and another corresponding to the universe of harassed people – called "H". Let us exemplify what these two columns mean in the case of white professors. Although the white professors who suffered harassment correspond to 94.8% (column H) of the total professors harassed, when normalized by the number of white professors who answered the questionnaire, 42.5% of them suffered moral harassment. This is an important distinction, because the amount of responses in each category and of characteristics is not the same. We also conducted ${\chi }^2$ tests with Cramer‘s V for effect size in each line, for example testing if the prevalence of reported moral harassment was higher among white professors, compared with white staff and white students and so on.

This table also discriminates these data in terms of sex designation at birth, gender identity, race/ethnicity, and sexual orientation. An analysis of the table allows us to conclude that harassment is not homogeneously distributed in all categories. In general, it is observed that moral harassment is more prevalent in women and black, bisexual, trans, and non-binary people. This trend is less evident in the case of students, where moral harassment has a more homogeneous distribution.

It is worth noting that the data can present an important statistic fluctuation because, although the total number of people who reported moral harassment is high, some percentages are very high due to the fluctuation of small numbers when subdivided into several categories. For example, in the case of technical administrative staff, there are 3 people who declared themselves to be non-binary and all of them reported having suffered moral harassment, which results in the 100% occurrence in column "R".

Regarding the statistically significant differences, they were found between the main groups, students reporting less moral harassment than professors and staff. In addition, we found differences by sex assigned at birth and gender identity (men as well as women). Differences were also encountered between the main groups among people declared white and black (with the largest effect size – black staff reporting more harassment), and among heterosexual, and bisexual people.

We also performed tests to verify if our results were statistically significant regarding the differences found in terms of gender, race and sexual orientation. For this tests we used all categories together (sample size is then 6051) and verified that women and other gender identities are more frequently victims of moral harassment than men ( ${\chi }^2$ (2, 6051) = 163.93, p $<$ 0.001, V=0.16 ). In terms of sexual orientation, homosexuals and bisexuals are more victims of moral harassment than heterosexuals ( ${\chi }^2$ (4, 6051) = 70.01, p $<$ 0.001, V=0.11 ). In terms of race and ethnicity, the differences found in the frequencies of victims of moral harassment are not statistically significant.

Table III has the same structure as Table II, but the data corresponds to people who have been victim of sexual harassment. The percentage of sexual harassment is around 12% of respondents in all categories.

Table III
The structure is the same as in Table II, but here is for Sexual harassment.

Sexual Harassment
	Professors		Staff		Students		${\chi }^2$	p	Cramer’s V
Total number	77		71		567
Percentage	10.4%		13.63%		11.83%		3.02	0.22	0.02
	R	H	R	H	R	H
Woman	14%	74%	18,2%	84,5%	16,5%	88%	2.49	0.29	0.03
Man	5,9%	26%	5,7%	15,5%	3,8%	12%	4.22	0.12	0.04
Woman	14,1%	74%	17,9%	83,1%	16,4%	86,4%	2.07	0.35	0.02
Man	5,9 %	26%	5,3%	14 %	3,5%	10,9%	4.94	0.08	0.05
Travesti	0%	0%	0%	0%	31,8%	0,35%	0.47	0.26	0.49
Trans Woman	0%	0%	10%	1,4%	24,3%	0,35%	2.25	0.13	0.50
Trans Man	0%	0%	0%	0%	18%	0,35%	-	-	-
Non-binary	0%	0%	3,2%	1,4%	19,3%	1,23%	0.33	0.09	0.57
Others	0%	0%	0%	0%	9%	0,35%	0.1	0.07	0.75
While	10,2%	92,9%	12,5%	82,8%	11,6%	79,6%	2.20	0.33	0.02
Black	0%	0%	18,5%	7,8%	10,5%	6,4%	2.54	0.28	0.08
Parda	11,5%	4,2%	12,2%	7,8%	13,5%	13%	0.17	0.92	0.02
Asian	34,5%	2,8%	0%	0%	4,2%	0,2%	5.03	0.08	0.40
Indigenous	0%	0%	33,3%	1,6%	28,6%	0,7%	0.03	0.87	0.04
Heterosexual	10,1%	88,3%	13,5%	90,1%	9,7%	58,9%	7.83	0.02	0.04
Homosexual	10,4%	6,5%	10%	2,8%	13,5%	8,6%	0.48	0.79	0.03
Bisexual	18,1%	5,2%	20%	5,6%	21,7%	29,6%	0.19	0.91	0.01
Asexual	0%	0%	0%	0%	11,7%	0,7%	0.53	0.47	0.18
Other	0%	0%	50%	1,4%	18,5%	2,12%	2.00	0.18	0.17

Besides being less prevalent than moral harassment, sexual harassment shares some similarities with it: it is not homogeneously distributed among all different profiles. It occurs much more frequently among women (around 15% of women who responded the survey reported having suffered sexual harassment against 5% of men) and it is around two times more frequent in bisexual than in heterosexual and homosexual people. The higher prevalence of sexual harassment among women and other minorities was already observed in previous studies, as in National Academies of Sciences et al. (2018), Anteneodo et al. (2020) and in references therein.

The problem of statistical fluctuation in this case is more relevant than in the case of moral harassment because the absolute number of responses is smaller. It is then important to realize that, although the percentage of indigenous and non-cis people (in case of students) who reported moral harassment is very high compared to other categories, they represent a small number of responses. It would be important to make an effort to improve this statistic to have a clearer picture of this discrimination.

Concerning the statistically significant differences using the ${\chi }^2$ test, they were only found, with a small effect size, among heterosexuals. Heterosexual staff declaring more sexual harassment than heterosexual professor than students. For all other subdivisions the differences between the tree categories were not statistically significant.

We finally performed tests to verify if our results were statistically significant regarding the differences found in terms of gender, race and sexual orientation in the case of victims of sexual harassment. The results are very similar to what we found for the case of moral harassment: women and other gender identities are more frequently victims of sexual harassment than men ( ${\chi }^2$ (2, 6051) = 208.55, p $<$ 0.001, V=0.16 ). Homosexuals and bisexuals are more frequently victims of sexual harassment than heterosexuals ( ${\chi }^2$ (4, 6051) = 94.42, p $<$ 0.001, V=0.12) and we did not detected statistical significance in terms of race and ethnicity for the victims of sexual harassment.

We also separated the responses by area of expertise, and we did not observe important differences in terms of percentage of harassment victims per area (figure not shown here).

Profile of the Harassers

For respondents who have suffered moral and/or sexual harassment, the questionnaire had an expansion to include some questions aiming to understand the profile of the aggressor. We had some options which are shown in the Fig. 1 on the left. We also differentiated the three categories, each represented with different symbols as indicated in the legend.

Figure 1
Percentage of people who were victims of moral (left) and sexual (right) harassment by a person indicated in the list on the left. Respondents were allowed to mark more than one option. Different categories are represented in different symbols, which are indicated in the legend.

From this figure, we can draw some observations. The first one concerns the gender difference in terms of type of harassment: sexual harassment is mostly committed by men (a male professor, a male colleague or a male from the staff), but moral harassment is also done by women (although in lower percentage compared to men). Previous studies have reported that men are more likely than women to commit sexual harassment (National Academies of Sciences et al. 2018).

We performed tests comparing if the responses between different categories were statistically significant. In the case of moral harassment, all were significantly different except the aggressor being a female professor (${\chi }^2$ (2, 2435) = 1.16, p = 0.56, V=0,.02) and from outside UFRGS (${\chi }^2$ (2, 2435) = 3.29, p = 0.19, V=0.04.). The values in parenthesis for the ${\chi }^2$ correspond to the degrees of freedom and sample size. Regarding sexual harassment they were found in the aggressor being a male colleague (${\chi }^2$ (2, 715) = 43.28, p = 0.03, V=0.1), a male staff (${\chi }^2$ (2, 715) = 43.28, p $<$ 0.001, V=0.25.) and a female staff (${\chi }^2$ (2, 715) = 20.93, p $<$ 0.001, V=0.17) and role not mentioned (${\chi }^2$ (2, 715) = 22.32, p $<$ 0.001, V=0.17.).

Another point is the difference between the categories: members of the staff suffer harassment from all other categories, while professors and students almost do not register "a male/female from the staff" as harasser. A male professor is the most common type of aggressor for students and is very high among staff as well. We would like to recall that the percentage of male professors is 53.5%. This gender gap is much smaller than the difference between the percentage of times that a male professor is cited as harasser and a female professor is cited as harasser, thus not explaining this observation.

On the harassment report

We also asked the victims of moral or sexual harassment if they reported the harassment and, if not, what are the main reasons for this.

Our data demonstrate that harassment is hardly ever reported, as detailed by category in Table IV. In both cases of harassment, students are the ones who report less, and these differences are statistically significant. We also observe that sexual harassment is less reported than its moral counterpart. No significant statistical differences were found between the groups regarding the frequency of reported sexual harassment.

Table IV
Percentage of victims of harassment who had reported it. The last three columns are ${\chi }^2$ tests, p-values, and Cramer’s V for effect size.

	Professors		Staff		Students		${\chi }^2$	p	Cramer’s V
Moral Harassment	12.7%		19.6%		7.5%		89.89	$<$ 0.001	0.09
Sexual Harassment	6.5%		11.3%		7.4%		4.60	0.33	0.02

To understand why most victims do not report harassment, we asked the reasons and gave them four multiple-choice options. These options are listed in Fig. 2 with the percentage of the responses to each one separated in three categories. In the case of sexual harassment, the most common reason to not report is the fact that the victim believes that they do not have proof or witness. In terms of moral harassment, almost 50% of students victim of moral harassment were afraid that the harasser interfere in the process.

Figure 2
Reasons for not reporting moral (left) and sexual (right) harassment are listed on the left. Professors, students, and technical administrative staff are discerned through the different symbols listed in the legend.

In moral harassment all differences were statistically significant among the tree categories except “I believed that the available channels would not proceed with my complain” (${\chi }^2$ (2, 5893) = 3.52, p = 0.17, V=0.04.). Regarding sexual harassment, we found significant statistical differences only in “I didn’t know to whom or where to resort” (${\chi }^2$ (2, 660) = 16.10, p$<$ 0.001, V=0.16.) and “I believed that the available channels would not proceed with my complain” (${\chi }^2$ (2, 660) = 7.33, p = 0.03, V=0.10).

Because male professors are the most common aggressors in this category, as shown in Fig. 1, this hierarchical relationship possibly creates more obstacles for the complaint. Other reasons are roughly equally registered as reasons to not denounce. The fear of denouncing is also found in other studies as (National Academies of Sciences et al. 2018), where they report that, for women, there is "an accurate perception they may experience retaliation or other negative outcomes associated with their personal and professional lives".

Also, when the harassment was notified, we asked what was the reporting channel used. We proposed four options of existing channels in the university, which are discriminated in Fig. 3, together with the percentage of each choice per category. As we can observe, it is very rare to use these options to report harassment, suggesting either discredit or unfamiliarity with these officials channels proposed by the university. Only among sexual harassment those differences were statistically significant (${\chi }^2$ (8, 232) = 34.46, p $<$ 0.001, V=0.42).

Figure 3
For respondents who reported moral harassment (left) and sexual harassment (right) there were four options to identify which were the reporting channels used to denounce the harassment. Professors, students, and technical administrative staff are discerned using different symbols as defined in the legend.

What is perceived as harassment

In this section we discuss which type of episodes are perceived as harassment. Unlike the previous sections of this article, where we differentiated the responses of professors, staff, and students, here we separate the responses by gender identities: we discern among respondents who self-declared as "women", "men", and "others".

Figure 4 shows how all the groups perceive the listed episodes as situations that motivate moral harassment. However, women tend to perceive most situations listed as harassment compared with other groups, with the exception of unwanted touches and unwanted sexual encounters that are most often perceived as bullying by men (regardless of the aggressor’s authority). This fact should be analyzed considering that 8.99% non-heterosexual men reported having been victims of some type of sexual harassment versus 2.86% heterosexual (${\chi }^2$ (2, 2261) = 34.35, p $<$ 0.001, V=0.12) and 45.39% of non-heterosexual men reported suffered any type of moral harassment versus 25.99% heterosexuals (${\chi }^2$ (2, 2261) = 64.30, p $<$ 0.001, V=0.17).

Figure 4
Percentage of students, professors, and staff who consider the above acts as moral harassment separated by gender and the authority of the harasser. Full text of the summarized items: ”Purposeful and systematic assignment of tasks inferior or superior to your competences”, ”Mock my origins, nationality, religious beliefs, or political beliefs”, and ”Be discredited in the presence of colleagues, superiors, or subordinates”.

All the differences between gender were statistically significant except threats by physical violence and physical assaults with (respectively ${\chi }^2$ (2, 6051) = 0.01, p = 0.91, V=0 and ${\chi }^2$ (2, 6051) = 0.22, p = 0.64, V=0.01) and without authority (respectively ${\chi }^2$ (2, 6051) = 5,34, p = 0,07, V=0,03, ${\chi }^2$ (2, 6051) = 1.67, p = 0.43, V=0.17). And, unwanted messages or phone calls, with authority ${\chi }^2$ (2, 6051) = 0.16, p = 0.69, V=0.05).

Figure 5 shows that there is a difference between the situations perceived as sexual harassment. Situations involving touching, inviting, and unwanted communication are often perceived as sexual harassment. In addition, more women and people with other gender identities perceive physical violence and threats of physical violence as sexual harassment than men. The greatest gender differences are in the situations most strongly perceived as sexual harassment by all groups, where women tend to perceive those situations as sexual harassment when compared with the other groups. The only difference that was not statistically different was "purposeful and systematic assignment of tasks inferior or superior to your competences in the contexts with authority" (${\chi }^2$ (2, 6051) = 5.52, p = 0.67, V=0.03).

Figure 5
Percentage of students, professors, and staff who consider the above acts as sexual harassment separated by gender and the authority of the harasser. Full text of the summarized items: "Purposeful and systematic assignment of tasks inferior or superior to your competences", "Mock my origins, nationality, religious beliefs, or political beliefs", and "Be discredited in the presence of colleagues, superiors, or subordinates".

In short, women and people with other gender identities are generally more likely to become victims of moral and sexual harassment than men. On the other hand, men, who are most often reported as harassers, are the least likely to perceive reported situations as sexual and moral harassment.

Which types of harassment are more common

Figure 6 shows that people who identified themselves with other gender identities (trans men and women) reported more episodes of moral harassment than men and women, except for unfair criticism of their work. In the same direction, women also suffered more episodes of moral harassment than men. Thus, there is a pattern of moral harassment directed at gender minorities.

Figure 6
Percentage of forms of moral harassment that students, professors, and staff claim to have suffered separated by gender. Full text of the summarized items: "Purposeful and systematic assignment of tasks inferior or superior to your competences", "Mock my origins, nationality, religious beliefs, or political beliefs", and "Be discredited in the presence of colleagues, superiors, or subordinates".

Among those difference are statistically significant sexist jokes (${\chi }^2$ (2, 6051) = 23.02, p $<$ 0.001, V=0.18), unwanted suggestive looks and gestures (${\chi }^2$ (2, 6051) = 29.27, p $<$ 0.001, V=0.20), unwanted invitations (${\chi }^2$ (2, 6051) =8.82, p = 0.03, V=0.11) with more prevalence among women and people with other gender identities. Were also statistically significant, with higher prevalence among people with other gender identities isolation from the rest of the group (${\chi }^2$ (2, 6051) = 8.82, p = 0.01, V=0.11), ignored presence (${\chi }^2$ (2. 6051) = 5.94, p = 0.05, V=0.09), physical threat (${\chi }^2$ (2. 6051) = 12.51, p = 0.02, V=0.13) and physical aggression (${\chi }^2$ (4, 6051) = 18.95, p $<$ 0.001, V=0.16).

Regarding the frequency of sexual harassment episodes suffered (Figure 7), once again people with other gender identities have a higher prevalence than women and men. We found significant statistical in discriminatory jokes (${\chi }^2$ (2, 6051) = 26.61, p $<$ 0.001, V=0.10), comments of sexual nature (${\chi }^2$ (2, 6051]) = 18.95, p $<$ 0.001, V=0.16), unwanted looks and gestures (${\chi }^2$ (2, 6051) = 23.55, p $<$ 0.001, V=0.16), unwanted invitations (${\chi }^2$ (2, 6051) = 18.95, p $<$ 0.001, V=0.1), unwanted touches (${\chi }^2$ (2, 6051) = 6.50, p =0.04, V=0.05), presence ignored (${\chi }^2$ (2, 6051) = 12.31, p =0.002, V=0.05), and physical threat (${\chi }^2$ (2, 6051) = 11.51, p = 0.003, V=0.07). Two points draw attention. Women suffer more with gestures and unwanted looks than the other two groups. In addition, men suffered more insistence of unwanted sexual encounters than other groups, and more unwanted invitations than women.

Figure 7
Percentage of forms of sexual harassment that students, professors, and staff claim to have suffered separated by gender. Full text of the summarized items: ”Purposeful and systematic assignment of tasks inferior or superior to your competences”, ”Mock my origins, nationality, religious beliefs, or political beliefs”, and ”Be discredited in the presence of colleagues, superiors, or subordinates”.

The literature has already pointed out more or less similar prevalence of heterosexual men and women reporting sexual coercion among university samples (Larimer et al. 1999, McConaghy & Zamir 1995). However, it should be noted that some authors have questioned the manner in which the question regarding coercion is asked and the impact on the interpretation of what coercion is between heterosexual men, which may be different from heterosexual women given the naturalization of harassment (Hogben et al. 2008). It is also important to note that among those who did not suffer any type of sexual harassment, 18.57% of men declared themselves to be non-heterosexual, while amid those who suffered this percentage rises to 43.96%. In this way, sexual harassment patterns also target gender (and sexual) minorities, reinforcing the logic of moral harassment (Braun et al. 2009).

It is important to realize that sexual harassment and discrimination are associated. The former is a way in which the latter can manifest, and it consists in three types of harassing behavior, according to (National Academies of Sciences et al. 2018): gender harassment, unwanted sexual attention, and sexual coercion. The report affirms that "The distinctions between the types of harassment are important, particularly because many people do not realize that gender harassment is a form of sexual harassment." Our results show that women classify typical examples of gender harassment (e.g. "Be discredited in the presence of colleagues", "Unfair criticism related to work") as moral harassment.

Discussion and Conclusions

In this paper we analyzed the answers of professors, staff, and students from UFRGS regarding sexual and moral harassment. The moral harassers are males and, in the majority of cases, professors, followed by male colleagues while victims are predominantly women, homosexual, and bisexual.

In the analysis of sexual harassment, again males are the perpetrators, but in this case male professors and colleagues share the prevalence. It is important to notice that even between colleagues some hierarchy is present, and, therefore, the power structure also plays a role in the event. Likewise in this case, women, homosexuals, and bisexuals are the victims. The larger impact on women is not surprising since this result was observed in previous studies (National Academies of Sciences et al. 2018). Some studies have identified that people of color are more frequently victims of harassment than white people (National Academies of Sciences et al. 2018). In this work we were not able to detect if black+parda are more often victims of harassment than white people because the differences we found between them were not statistically significant. It can be an effect of small sample size and it would be nice to promote this study in other institutions to obtain more data and evaluate this aspect.

One important element in these observations is the perception of what count as moral and/or sexual harassment. In the case of moral harassment, both men and women include a wide spectrum of attitudes in their description. However, there are important differences between the perception of men and women. A larger percentage of men considers unwanted touch as moral harassment (when in principle this should be considered as sexual harassment) when compared to women, and a lower percentage of men, compared to women, considers discriminatory jokes as moral harassment. Complementary, women consider unwanted invitations as sexual harassment more than men. These differences in perceptions were observed in previous studies (Studzikska 2015). Some studies even identify that, for some people, women are responsible for being harassed and even enjoy being provoked (Banerjee & Banerjee 2011). In this sense, it is fundamental to foment a broad understanding of what harassment is in order for the harasser to know what the limit of their action should be, and likewise for the victim to know what attitude should be institutionally reported if it occurs. Our results indicate that this awareness still needs to be generated.

In addition to not having a clear definition of what is moral or sexual harassment, many respondents do not trust the institutions to solve the problem. Since harassment involves power, the victim is afraid of retaliation and humiliation. One of the potential reasons for distrust in the organizations is that they are based in a male-dominated structure which creates the environment for the harassment to persist (Freitas 2001). This is a worrying result, mainly because previous studies have identified that one of the characteristics of environments which are most associated with higher rates of sexual harassment is an organizational environment that communicates tolerance of sexual harassment (National Academies of Sciences et al. 2018).

To face this problem some actions need to be done at very different levels in the university (Clancy et al. 2020, Tenbrunsel et al. 2019). It is important to develop mechanisms to ensure gender and race equality in academia. Sexual and moral harassment need to be framed not only as a legal issue, but also as an ethical problem. Besides the necessity of creating strategies to prevent harassment and for victim support, it is also crucial to develop transparency strategies for harassment cases in the university: how many they are, how they are being analyzed, and how perpetrators are being punished.

ACKNOWLEDGMENTS

We thank the members of the HeForShe UFRGS commission for the discussion and critical reading of these results and CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) for partially financing this study.

• 1
  There is no good translation for travesti to English, once it is a specific female identity construction of Brazil and other Latin American countries, distinct from transvestite (the erroneous English translation) or transgender/transsexual.
• 2
  In 2019, the university signed the term of adhesion to the regional Gaúcho Committee for the HeForShe initiative, in which it has participated since 2017. This research was the first initiative of the local work group.

REFERENCES

• ANDRADE CB & ASSIS SG. 2018. Assédio moral no trabalho, gênero, raça e poder: revisão de literatura. Rev Bras Saude Ocup 43: e11.
• ANTENEODO C, BRITO C, ALVES-BRITO A, ALEXANDRE SS, D’AVILA BN & MENEZES DP. 2020. Brazilian physicists community diversity, equity, and inclusion: A first diagnostic. Phys Rev Phys Educ Res 16: 010136.
• ATHERNON TJ, BARTHELEMY RS, DECONINCK W, FALK ML, GARMON S, LONG E, PLISCH M, SIMMONS EH & REEVES K. 2016. LGBT Climate in Physics: Building an Inclusive Community. Am Phy Soc, College Park, MD. URL https://www.aps.org/programs/lgbt/upload/ LGBTClimateinPhysicsReport.pdf
  » https://www.aps.org/programs/lgbt/upload/ LGBTClimateinPhysicsReport.pdf
• BANERJEE A & SHARMA B. 2011. Gender differences in perception of workplace sexual harassment among future professionals, Ind Psychiatry J 20: 21.
• BRAUN V, SCHMIDT J, GAVEY N & FENAUGHTY J. 2009. Sexual coercion among gay and bisexual men in Aotearoa/New Zealand. J Homosex 56: 336-360.
• CARAN VCS, SECCO IAO, BARBOSA DA & ROBAZI MLCC. 2010. Assédio moral entre docentes de instituição pública de ensino superior do Brasil. Acta Paul Enferm 23: 737-744.
• CLANCY K, CORTINA LM & KIRKLAND AL. 2020. Use science to stop sexual harassment in higher education Proc Natl Acad Sci USA 117: 22614.
• COGENLI MZ & BARLI O. 2013. The Exposure of Psychological Violence (Mobbing) in Universities and an Application to the Academicians. Procedia Soc Behav Sci 93: 1174-1178.
• FREITAS ME. 2001, Assédio moral e assédio sexual: faces do poder perverso nas organizações. Rev Admin Empr 41: 8.
• GALE S, MORDUKHOVICH I, NEWLAN S & MCNEELY E. 2019. The Impact of Workplace Harassment on Health in a Working Cohort. Front Psychol 24: 1181.
• GUIMARÃES CA, CANÇADO VL & LIMA RJC. 2016. Workplace moral harassment and its consequences: A case study in a federal higher education institution. Rev Adm 51: 151-164.
• HIRIGOYEN M-F. 1998. Le harcèlement moral: la violence perverse au quotidien, Paris: Syros, 214 p.
• HOGBEN M, BYRNE D & HAMBURGER ME. 2008. Coercive heterosexual sexuality in dating relationships of college students: Implications of differential male-female experiences. J Psychol Human Sex 8: 69-78.
• IBGE. 2020. Conheça o Brasil - População Cor e Raça. URL https://educa.ibge.gov.br/jovens/conheca-o-brasil/populacao/18319-cor-ou-raca.html
  » https://educa.ibge.gov.br/jovens/conheca-o-brasil/populacao/18319-cor-ou-raca.html
• INSTITUTO HUMANITAS UNISINOS. URL encurtador.com.br/FLT23
  » encurtador.com.br/FLT23
• KENIG S & RYAN J. 1986. Sex Differences in Levels of Tolerance and Attribution of Blame for Sexual Harassment on a University Campus. Sex Roles 15: 535.
• LARIMER ME, LYDUM AR, ANDERSON BK & LARIMER APT. 1999. Male and female recipients of unwanted sexual contact in a college student sample: Prevalence rates, alcohol use, and depression symptom. Sex Roles 40: 295-308.
• LEYMANN H. 1990. Mobbing and Psychological Terror at Workplaces. Violence Vict 5: 119-126.
• MARTIN-STOREY AA, PAQUETTE G, BERGERON M, DION J, DAIGNEAULT I, HÉBERT M & RICCI S. 2018. Sexual Violence on Campus: Differences Across Gender and Sexual Minority Status. J Adolesc Health 62: 701.
• MCCONAGHY N & ZAMIR RR. 1995. Heterosexual and homosexual coercion, sexual orientation and sexual roles in medical students. Arch SeX Behav 24: 489-502.
• NATIONAL ACADEMIES OF SCIENCES, ENGINEERING, AND MEDICINE. 2018. Sexual Harassment of Women: Climate, Culture, and Consequences in Academic Sciences, Engineering, and Medicine; Washington, DC, The National Academies Press, 331 p.
• ROSA BS, BARBOSA MC, PAVANI D, BRANDELLI A, NARDI HC & BRITO C. 2020. Pesquisa sobre percepção de assédio moral e sexual relativo à gênero na UFRGS. URL www.ufrgs.br/meninasnaciencia/pesquisa-assedio-na-ufrgs
  » www.ufrgs.br/meninasnaciencia/pesquisa-assedio-na-ufrgs
• SILVEIRA RS, NARDI H C & SPINDER G. 2014. Articulações entre gênero e raça/cor em situações de violência de gênero. Psicol Soc 26: 323.
• STUDZIKSKA A. 2015. Gender Differences in Perception of Sexual Harassment Doctoral Thesis. University Toulouse. URL https://tel.archives-ouvertes.fr/tel-01320713/ document.
• TENBRUNSEL AE, MCKENZIE RR & DIEKMANN KA. 2019. Sexual Harassment in Academia: Ethical Climates and Bounded Ethicality. Annu Rev Psyc 70: 245.
• TWENGE JM, SHERMAN RA & WELLS BE. 2016. Changes in American Adults’ Reported Same-Sex Sexual Experiences and Attitudes, 1973-2014. Arch Sex Behav 45: 1713.
• VERBA S, DINUNZIO J & C SPAULDING C. 1983. Harvard Sexual Harassment Survey. doi:https://doi.org/10.7910/DVN/KWJLIP
  » https://doi.org/10.7910/DVN/KWJLIP
• ZOTARELI V, FAUDES A, OSIS MJD, DUARTE GA & DE SOUZA MH. 2012. Violência de gênero e sexual entre alunos de uma universidade brasileira. 2012. Rev Bras Saúde Mater Infant 12: 37.

Publication Dates

• Publication in this collection
  08 Apr 2022
• Date of issue
  2022

History

• Received
  20 Nov 2020
• Accepted
  23 Mar 2021