Exposure to Workplace Sexual Harassment among Women and Men Farmworkers in the U.S. and Mexico

Abstract

This study explores experiences relevant to workplace sexual harassment (WSH) in agriculture among men and women farmworkers in California (U.S.) and Michoacán (Mexico). Anecdotal evidence documents women farmworkers having to endure behavioral, verbal, and physical WSH including sexual ogling, degrading language, groping, and requests for sex in exchange for work. We include survey comparisons between men and women in California and Michoacan on WSH among farmworkers. We conducted 197 farmworker surveys (38 men and 59 women in California; 40 men and 60 women in Michoacán). Community advisory boards contributed expertise and input for study strategies, materials, and dissemination. Survey participant ages ranged from 23 to 54 years old. Half worked in Mexico, 68% were married, 80% had children, and 47% had less than 7 years of education. Most farmworkers spoke Spanish and Purhépecha, an indigenous language spoken by the Purhépecha people in Michoacán. We used two strategies to measure WSH exposure in the previous year: (1) direct inquiry-based Survey items (asking “Have you ever been the victim of or bystander to workplace sexual harassment? ”) documenting WSH among women (49%) and men (21%) in California and among women (7%) and men (13%) in Michoacán, and (2) behavior-based WSH items (using explicit examples of WSH behaviors perpetrated against the participant or witnessed by the participant as a bystander) documenting WSH among women (as high as 53%) and men (as high as 45%) in California and among women (as high as 65%) and men (as high as 68%) in Michoacán. Women farmworkers in California reported WSH experiences exceeding those of men. Reported WSH experiences in Michoacán were similar for men and women. Farmworkers identified WSH perpetrators as coworkers more than leadership. The frequency of exposure ranged from daily, weekly, monthly, and up to multiple times a year. Of 46 direct inquiry-based WSH incidents, only one perpetrator was punished, and at least half of all victims said they were forced to change their jobs. The findings of this study inform the development of WSH prevention efforts, such as education tools, support for efforts to facilitate reporting, protections against retaliation for workers, and promoting accountability for perpetrators. This information supports the promotion of policy recommendations and preventive approaches for WSH.

Workplace sexual harassment (WSH) is a ubiquitous phenomenon across age, race, marital status, and occupation (Fitzgerald et al., 1988; Gruber and Bjorn, 1982; Ménard et al., 2003). The Equal Employment Opportunity Commission (EEOC) guidelines state that harassment can include sexual harassment, e.g., unwelcome sexual advances, requests for sexual favors, and other verbal or physical harassment of a sexual nature. Harassment can also include offensive remarks relating to a person’s sex. Both the victim and the harasser can be a woman or a man, and the victim and harasser can be of the same sex. Harassment is illegal when it is so frequent or severe that it creates a hostile or offensive work environment or when it results in an adverse employment decision (such as the victim being fired or demoted). The harasser can be the victim’s supervisor, a supervisor in another area, a coworker, or someone who is not an employee of the same employer, such as a client or customer.

Although perpetrators and victims can be of either sex, women are at high risk of victimization (Kim et al., 2016; Prado et al., 2020). Approximately 35% to 50% of women are sexually harassed at some point in their working lives (Gutek and Done, 2001). In the U.S., 50% of all working women (Fitzgerald and Cortina, 2018), 40% to 55% of university students and faculty (Moylan and Wood, 2016; Richman et al., 1999), and 41% to 60% of women soldiers (Barth et al., 2016; Gurung et al., 2018) experience WSH. Among working-class Latinas, up to 60% report WSH (Cortina, 2001).

The agricultural industry differs dramatically from typical middle-class employment, and its social structures help perpetuate WSH (Flores, 2018). Accordingly, the prevalence of WSH among women farmworkers is higher than in most industries in the United States. Circumstances such as reduced worker protections, field work, warehouse work, wage theft, unpredictable hours, and male-dominated crews increase the risk of this kind of exploitation among women agricultural workers. Of women farmworkers surveyed in California, 80% experienced WSH (Waugh, 2010). Despite its ubiquity in the agricultural workplace, consumers are largely oblivious to the suffering endured by farmworkers, and this invisibility perpetuates a cycle of exploitation of those whose labor sustains the food economy (Flores, 2018). Lastly, migrant women’s experiences of WSH are compounded by legal and cultural factors that further marginalize them in the workplace (Villegas, 2019).

Workplace Sexual Harassment and Gender in the U.S. and Mexico

Focus groups and one survey reveal that WSH against women in agriculture is widespread in the U.S. (Kim et al., 2016; Murphy et al., 2015; Prado et al., 2020; Waugh, 2010) and Mexico (Andrade-Rubio, 2016; Arellano Gálvez, 2014; Prado et al., 2020). Farmworkers have noted that “just being female in the fields creates risks” (Castañeda and Zavella, 2003). The survey among women farmworkers in California revealed that 97% reported gender harassment (generalized sexist comments and behavior that conveyed insulting, degrading, and sexist attitudes), 53% reported unwanted sexual attention (ranging from unwanted, inappropriate, and offensive physical or verbal sexual advances to gross sexual imposition, assault, or rape), and 24% reported sexual coercion (the solicitation or coercion of sexual favors by promise of reward or threat of punishment) (Waugh, 2010).

The regulation and illegality of WSH in the U.S. differs from that in Mexico. For example, the U.S. has declared WSH illegal in all 50 states under the Civil Rights Act of 1964. It was also made illegal under court rulings defining “quid pro quo” and “hostile workplace” in Williams v. Saxbe (1976) and Meritor Savings Bank v. Vinson (1986). The EEOC is the federal agency responsible for prosecuting WSH in the United States. In Mexico, WSH was acknowledged in the 1991 federal penal code. It is considered a crime in 16 of 32 Mexico states. It was also made illegal under the 2007 Ley General de Acceso de las Mujeres a una Vida Libre de Violencia (General Law of Access for Women to Live a Life Free of Violence). Mexico does not have an EEOC-equivalent regulatory agency to prosecute and prevent WSH. Additionally, women farmworkers in Mexico face greatly reduced occupational, social, and cultural support. For example, women farmworkers in Mexico may face severe WSH, domestic violence in the home, and are often blamed for the WSH perpetrated against them. In many cases, women are expected to endure the oppression of men because of the widespread cultural assumption that men have greater authority and that men are superior (Andrade-Rubio, 2016; Arellano Gálvez, 2014; Calvario Parra, 2016; Zúñiga-Elizalde, 2008).

There are few data comparing the WSH experiences of men and women in the U.S. and Mexico, despite the fact that Mexico is the country that most frequently sends farmworkers to California and other regions in the U.S. (Hernandez et al., 2016). Separate focus groups among men and women farmworkers in California and Michoacán revealed that women experienced more frequent and severe WSH compared to men (Prado et al., 2020). Focus groups in Mexico among women farmworkers described a strong patriarchal system in which men may act with a sense of entitlement and express themselves in hypermasculine and violent ways (Andrade-Rubio, 2016). Although employers in the U.S. must abide by laws protecting workers against discrimination based on sex, employers in Mexico operate under a comparatively looser regulatory framework in which there is lack of follow-up, supervision, and less enforcement of the law (Andrade-Rubio, 2016; Arellano Gálvez, 2014; Prado et al., 2020).

This Study

We report here the results of a cross-sectional survey to investigate the WSH experiences among women and men farmworkers in California (U.S.) and Michoacán (Mexico). Comparing across genders and across countries broadens the impact of findings. Mexico is the country that most frequently sends farmworkers to California and the rest of the U.S. (Hernandez and Gabbard, 2019; Li et al., 2020). The state of Michoacán ranks third nationally in migrants from Mexico to the U.S., and most travel primarily to California, Texas, and Illinois (Deeb-Sossa, 2019); 47.6% of Michoacán migrants are in California (Li et al., 2020). For these reasons, we sampled with the aim of comparing similarities and differences between men and women farmworkers in California and Michoacán. We also offer a transcultural comparison to assess the similarities or diversities occurring in the two cultures (Valentín et al., 2005). Our aim is to inform and support the agricultural community, educators, researchers, and organizations working to prevent and respond to WSH.

Methods

Participants

Bilingual (English and Spanish or Spanish and Purhépecha) researchers and community members assisted in recruiting farmworkers from agricultural communities in Michoacán (40 men and 60 women) and California (38 men and 59 women). The Michoacán farmworkers spoke Spanish or Purhépecha, an indigenous language spoken by the Purhépecha people centered in the northwest region of Michoacán. Farmworkers were eligible to participate if they were at least 18 years old, currently employed in agriculture, and had worked for at least the past year (12 months) in agriculture.

Community Advisory Board

Study teams in Michoacán and California each convened local community advisory boards (CABs, or mesas consultivas). These included farmworkers, farmworker welfare advocates, non-profit representatives, legal entities, academics, industry officers, and community leaders. The CAB members assisted in the development of study strategies, materials, and plans for dissemination.

Sampling Sites

We sampled in California (U.S.) and Michoacán (Mexico). Researchers engaged with local community leaders and traveled to farmworker communities to recruit participants in their homes. Farmworkers in California were recruited with the help of local farmworker coalitions and migrant housing centers. Farmworkers in Michoacán were recruited with the help of local community leaders.

Survey Development

Focus group results (Prado et al., 2020), pilot work, and CAB input informed the survey development. Previous focus group results had revealed novel anecdotal evidence of men farmworkers being sexually harassed and women farmworkers perpetrating WSH (Prado et al., 2020). To document this novel mixed-gender perpetration of WSH in our quantitative efforts, we included items that collected information on WSH with different gender combinations of person A targeting person B (men on men, men on women, women on women, and women on men WSH perpetration). Following CAB review, we partnered with a local farmworker health clinic to pilot the survey with four women and two men farmworkers.

Data Collection

Farmworker survey recruitment efforts began in June and July of 2017. California ended recruitment in October 2017, and Michoacán ended recruitment in February 2018. We used convenience and purposeful sampling (Valerio et al., 2016). In California, participants were reached through door-to-door canvassing, flyers, and visits in popular areas such as laundromats, worship centers, and the administrative office at a migrant housing center in northern California. We left flyers with a phone number for participants to call and set up individual appointments to provide informed consent and complete the interviewer-administered survey in a private room away from employers and coworkers. Interviews in Michoacán were conducted in participants’ homes to avoid interaction with coworkers or employers. Most interviews were conducted in Spanish by our research interviewer, and a translator was used for Purhépecha speakers. All interviews lasted approximately one hour. CABs helped researchers choose participant incentives that were appropriate to their locale. California participants received a $50 gift card for a local department store, and Michoacán participants received a luggage bag worth $25. All study activities were approved by the Institutional Review Board at the University of California Davis (UCD IRB No. 946036-6).

Demographics, Occupational Characteristics, and Work Infrastructure

We collected demographics (age, schooling, birthplace, preferred language, and other related variables) and occupational characteristics (job description and tasks, piecework versus hourly payment, presence of family members in the work crew). Questions on workplace infrastructure explored whether participants could identify a supervisor or management officer for reporting incidents of WSH. Workplace infrastructure questions also related to reporting WSH, if a solution to WSH was presented, if retaliation had occurred, and if the perpetrator had been punished.

Sexual Harassment Experiences

We used two methods to collect information in WSH. The first method included direct-inquiry based items (asking “Have you ever been the victim of or bystander to workplace sexual harassment?” and answering yes or no), and the second was behavior-based items (using explicit examples of WSH behaviors perpetrated against the participant or witnessed by the participant as bystander and answering never, sometimes, frequently, very frequently, or always) (Chan et al., 2008; Cortina, 2001). Extensive research (Fitzgerald et al., 1995) has documented the psychological dimensions, or types, of sexual harassment. The behavior-based items were adapted from the psychometrically rigorous Spanish SEQ-L (Sexual Experience Questionnaire-Latina), which is a validated instrument for measuring sexual harassment among Latinas (Cortina, 2001). The SEQ-L addresses three distinct types of sexual harassment that we used in our survey: sexual hostility (behaviors involving sexually offensive remarks and comments), unwanted sexual attention (e.g., ogling, touching, requests for dates), and sexist hostility (misogynist comments without sexual content).

A fourth type of behavior-based WSH was included to capture incidents described as quid pro quo (coercion, offer of favorable work conditions in return for sexual favors). This type of WSH was informed by our CAB members, farmworkers in our focus groups, previous literature describing these experiences among farmworker women (Waugh, 2010), and adaptations from validated surveys (Fitzgerald et al., 1995; Fitzgerald and Hesson-McInnis, 1989; Stark et al., 2002). The response options for the quid pro quo survey items were the same as for our other behavior-based WSH types.

Our survey aim was to document both direct and indirect WSH. Indirect WSH incidents included experiences that the participant witnessed, was close enough to hear and see, but was not involved in, i.e., was a bystander. Direct WSH incidents were those in which the participant was targeted. The survey items were formatted to record both direct and indirect WSH (table 1). Our study combined direct and indirect WSH as a single summary variable, consistent with the realities of WSH in agricultural work crews, where WSH incidents affect both the target and bystanders (Prado et al., 2020).

Table 1.

Workplace sexual harassment (WSH) types and their associated behavior-based survey items used in a survey of 197 Hispanic farmworkers in California (U.S.) and Michoacán (Mexico).^[a]

WSH Type[b] and Survey Item		Behavior-Based Inquiry
Gender discrimination (α = 0.835)
	GD1	Someone at your workplace makes comments that insult women, for example, saying: “Women do not make good supervisors” or something similar.
	GD2	Someone in your job makes crude and sexual jokes, stories, or comments that describe women negatively.
	GD3	Someone at your workplace says words to insult women, for example, saying they are “easy.”
Hostile behavior (± = 0.867)
	HB1	Someone at your workplace makes crude and sexual comments in front of other people.
	HB2	Someone at your workplace says offensive things to another person relating to their body or sex life.
	HB3	Someone at your workplace tries to make someone else speak about sexual things.
Unwanted sexual attention (± = 0.740)
	USA1	Someone at your workplace tries to talk to other workers about their sex life or sexual preferences, making those people uncomfortable.
	USA2	Someone at your workplace insists on having a sexual or romantic relationship with another worker even though that worker has told them that they didn’t want to.
	USA3	Someone at your workplace looks slowly at the body of another worker in a way that makes them feel uncomfortable or dirty.
	USA4	Someone at your workplace yells catcalls and whistles at another worker in front of other people in a way that makes them feel uncomfortable.
Quid pro quo (± = 0.886)
	QPQ1	Someone at your workplace has been offered better working conditions or a promotion in exchange for sexual favors.
	QPQ2	Someone at your workplace has been threatened or punished for not having accepted sexual advances.
	QPQ3	Someone at your workplace has suffered negative consequences for not having agreed to participate in sexual conduct.

^[a]Contact the lead author (kyprado@ucdavis.edu) for a copy of the survey format used in data collection.

^[b]Cronbach’s alpha (α) is provided for the total population.

Our four WSH types for the behavior-based survey items were gender discrimination (GD, characterized by sexist hostility), hostile sexual behavior (HB), unwanted sexual attention (USA), and quid pro quo (QPQ). Each WSH type had at least three behavior-based survey items. We asked the participants to report their experiences on a five-point frequency scale consisting of never (0 times), sometimes (1 to 3 times per year), frequently (4 to 6 times per year), very frequently (1 to 2 times per month), and always (1 to 2 times per week). The responses were later dichotomized (occurred vs. never occurred) for logistic regression analysis. Participants also indicated the perpetrator’s job position as coworker, contractor, transporter, crew leader, supervisor, or field owner. A summary leader variable was created by combining leadership positions (contractor, transporter, crew leader, supervisor, or field owner) into a single category of positions with authority. Participants were given the choice to decline to answer. Declined answers were considered missing values.

Survey Reliability

The survey items under all WSH types were tested for internal consistency and reliability using principal component analysis and calculating Cronbach’s alpha (α). Instrument subscales included men in Michoacán, women in Michoacán, the total sample in Michoacán, men in California, women in California, and the total sample in California. Responses were analyzed according to a five-point Likert scale. Final coefficient alphas for the subscales and the total 197 participants all measured above 0.7 (table 1). These results helped confirm that the SEQ-L items previously validated among women farmworkers also worked well among men farmworkers in California and farmworkers in Michoacán. This also helped us ensure that the quid pro quo (QPQ) items were consistent.

Analysis

Descriptive Analysis

Counts, percentages, and medians were calculated to characterize demographics, occupational characteristics, workplace infrastructure, WSH experiences, WSH frequency, perpetrator type, and mixed-gender WSH. We stratified measures by country and gender.

Statistical Analysis

Summary statistics, chi-squared, and Fisher’s exact tests were calculated for workplace infrastructure and WSH experiences. Summary statistics and Kruskal-Wallis testing was conducted on WSH frequencies. Statistical testing was conducted in SAS (ver. 9.4, SAS Institute, Cary, N.C.). These results were all evaluated against country and gender. An alpha value of α < 0.05 was considered statistically significant. Odds ratios included 95% confidence intervals. Preliminary results were presented to CAB members for review.

Results

Demographics, Occupational Characteristics, and Workplace Infrastructure

Demographic variables for the farmworkers were generally similar across country and gender (table 2). Most participants in California and Michoacán were born in Mexico. California men were about a decade older than California women and Michoacán men and women. California participants had the highest levels of education, especially women, of whom over half completed at least 12 years of education. Men were more likely to be married than women in both California and Michoacán. Women farmworkers in Michoacán were the only group in which the majority was not married. About a third were living with their partner, and about a third were living single. Most farmworkers spoke Spanish, and Purhépecha was spoken only in Michoacán (14 men and 3 women).

Table 2.

Selected demographics of 197 Hispanic farmworkers evaluated for workplace sexual harassment (WSH) in California and Michoacán. Values are numbers (percentages in parentheses).

Demographics	California (U.S.)		Michoacán (Mexico)
	Women (N = 59)	Men (N = 38)	Women (N = 60)	Men (N = 40)
Country of birth
Mexico	47 (80)	33 (87)	58 (97)	40 (100)
U.S.	12 (20)	5 (13)	2 (3)	0 (0)
Age (years)
Median	34	42.5	30.5	31.5
Interquartile range (IQR)	27 to 53	32 to 54	24 to 44	23.5 to 41
Education
0 to 3 years	7 (12)	4 (11)	8 (13)	9 (23)
4 to 7 years	10 (17)	15 (39)	21 (35)	16 (40)
8 to 11 years	11 (19)	6 (16)	28 (47)	13 (33)
12 years, trade, college	31 (53)	13 (34)	3 (5)	1 (3)
Have children
≥1	46 (78)	34 (89)	49 (82)	30 (75)
Civil status
Married	46 (78)	36 (95)	21 (35)	31 (78)
Single	12 (20)	1 (3)	17 (28)	5 (13)
Living with partner	1 (2)	0 (0)	20 (33)	3 (8)
Widow	0 (0)	1 (3)	1 (2)	1 (3)
Primary language
Spanish	58 (98)	37 (97)	57 (95)	26 (65)
English	1 (2)	0 (0)	0 (0)	0 (0)
Purhépecha	0 (0)	0 (0)	3 (5)	14 (35)

Occupational characteristic variables for the farmworkers differed by country and gender (table 3). About the same proportion of farmworkers in Michoacán and California moved with crops to find employment. However, farmworkers in California more commonly reported working seasonally than did Michoacán participants. Most farmworkers in California reported payment by the hour, whereas most Michoacán farmworkers reported payment by the day, crate, or week. Most farmworkers in Michoacán worked among family, and about half did so in California.

Table 3.

Selected occupational characteristics among 197 Hispanic farmworkers in California and Michoacán. Values are numbers (percentages in parentheses). Asterisks (*) indicate that women are significantly different (p < 0.05) from men within the same country.

Occupational Characteristics	California (U.S.)		Michoacán (Mexico)
	Women (N = 59)	Men (N = 38)	Women (N = 60)	Men (N = 40)
Movement follows crops	38 (64)*	14 (37)	41 (68)*	29 (73)
Seasonal work (not year-round)	57 (97)*	38 (100)	18 (30)	12 (30)
Payment scheme
Contract	1 (2)	1 (3)	6 (10)	1 (3)
Hour	58 (98)*	37 (97)	1 (2)	0 (0)
Piece	0 (0)	0 (0)	2 (3)	1 (3)
Other: day, crate, week	0 (0)	0 (0)	51 (85)*	38 (95)
Work among family	34 (58)*	17 (45)	49 (82)*	30 (75)

Workplace Infrastructure for Sexual Harassment with Direct-Inquiry Based Survey Items

Workplace infrastructure differed by country and gender (table 4). Most farmworkers knew of someone at work who could receive WSH reports. Being the target or witness of a WSH incident (table 4a) was more common in the U.S. (49% of women, 21% of men) than in Michoacán (7% of women, 13% of men). Among participants who experienced WSH in direct inquiry (table 4b), WSH was reported to leadership in 50% of the specific incidents in California across genders, compared to 20% of Michoacán incidents across genders. California incidents were reportedly resolved more frequently than those in Michoacán. No perpetrator was punished in California, and only one male farmworker in Michoacán reported the perpetrator being punished. Retaliation against the victim was present in over 90% of incidents reported in California and in at least 50% of incidents reported in Michoacán. The victims were forced to change their job in over 80% of incidents reported in California and in approximately 50% of incidents reported in Michoacán.

Table 4.

Selected workplace infrastructure characteristics related to sexual harassment among 197 Hispanic farmworkers in California (U.S.) and Michoacán (Mexico). Values are numbers (percentages in parentheses). Asterisks (*) indicate that women are significantly different (p < 0.05) from men within the same country.

	Workplace infrastructure characteristics for the total of 197 farmworkers	California (N = 97)		Michoacán (N = 100)
(a)		Women (N = 59)	Men (N = 38)	Women (N = 60)	Men (N = 40)
	Existence of designated individual to receive WSH reports	41 (69)*	37 (97)	54 (90)*	28 (70)
	Supervisors help available	40 (68)*	34 (89)	38 (63)*	15 (38)
	No help available	3 (5)	0 (0)	4 (7)	8 (20)
	Comfortable seeking supervisor	31 (53)*	37 (97)	50 (83)	31 (78)
	Not comfortable seeking help	9 (15)	1 (3)	1 (2)	1 (3)
	Direct inquiry if participant was target of WSH or bystander to WSH	29 (49)*	8 (21)	4 (7)	5 (13)
	Workplace infrastructure characteristics for the 46 incidents in direct inquiry	California (N = 37)		Michoacán (N = 9)
(b)		Women (N = 29)	Men (N = 8)	Women (N = 4)	Men (N = 5)
	Victim reported specific incident	16 (55)	4 (50)	1 (25)	1 (20)
	Solution was offered	19 (66)	4 (50)	1 (25)	1 (20)
	Perpetrator was punished	0 (0)	0 (0)	0 (0)	1 (20)
	Perpetrator changed crews	20 (69)	5 (63)	2 (50)	3 (60)
	Victim changed crews	22 (76)	7 (88)	1 (25)	2 (40)
	Victim suffered retaliation	28 (97)	8 (100)	2 (50)	3 (60)
	Victim forced to change their job	28 (97)	7 (88)	2 (50)	3 (60)

Sexual Harassment Experiences by Country and Gender with Behavior-Based Survey Items

WSH experience variables for farmworkers were stratified by country and gender (table 5). The WSH types in our survey included gender discrimination (GD), hostile behavior (HB), unwanted sexual attention (USA), and quid pro quo (QPQ). All four WSH types were experienced (table 5). California women farmworkers reported quid pro quo (7%) and unwanted sexual attention (20% to 53%). California men’s experiences included quid pro quo (3% to 5%) and hostile behavior (26% to 45%). Michoacán women reported exposure to quid pro quo (10% to 12%) and hostile behavior (25% to 65%). Michoacán men’s experiences included quid pro quo (5% to 10%) and gender discrimination (50% to 68%).

Table 5.

Selected experiences of workplace sexual harassment (WSH) and associations with women gender by country among 197 Hispanic farmworkers in California (U.S.) and Michoacán (Mexico). Values are numbers (percentages in parentheses). Asterisks (*) indicate that women are significantly different (p < 0.05) from men within the same country (p-values are based on chi-squared tests of association or Fisher’s exact test when cell size is <5).

	California (U.S.)			Michoacán (Mexico)
Survey Item[a]	Women (N = 59)	Men (N = 38)	OR (95% CI)	Women (N = 60)	Men (N = 40)	OR (95% CI)
GD1	26 (44)	14 (37)	1.35 (0.58, 3.12)	34 (57)	25 (63)	0.78 (0.35, 1.78)
GD2	25 (42)	16 (42)	1.01 (0.44, 2.31)	26 (43)*	27 (68)	0.38 (0.16, 0.88)
GD3	23 (39)	13 (34)	1.23 (0.52, 2.87)	34 (57)	20 (50)	1.31 (0.58, 2.92)
HB1	24 (41)	17 (45)	0.85 (0.37, 1.9)	39 (65)	27 (68)	0.89 (0.38, 2.09)
HB2	23 (39)	11 (29)	1.57 (0.65, 3.76)	30 (50)	23 (58)	0.74 (0.33, 1.65)
HB3	17 (29)	10 (26)	1.13 (0.45, 2.83)	15 (25)	11 (28)	0.88 (0.35, 2.18)
USA1	24 (41)*	6 (16)	3.66 (1.32, 10.09)	18 (30)	13 (33)	0.89 (0.38, 2.11)
USA2	12 (20)	2 (5)	4.59 (0.97, 21.84)	20 (33)	9 (23)	1.72 (0.69, 4.30)
USA3	31 (53)	12 (32)	2.40 (1.02, 5.63)	33 (55)	16 (40)	1.83 (0.81, 4.13)
USA4	19 (32)	11 (29)	1.16 (0.48, 2.83)	32 (53)	27 (68)	0.55 (0.24, 1.27)
QPQ1	4 (7)	1 (3)	2.69 (0.29, 25.04)	7 (12)	2 (5)	2.51 (0.49,12.75)
QPQ2	4 (7)	2 (5)	1.31 (0.23, 7.52)	6 (10)	4 (10)	1.00 (0.26, 3.79)
QPQ3	4 (7)	1 (3)	2.69 (0.29, 25.04)	6 (10)	3 (8)	1.37 (0.32, 5.83)

^[a]GD = gender discrimination, HB = hostile behavior, USA = unwanted sexual attention, QPQ = quid pro quo.

Few significant differences were found between men and women. Michoacán women experienced significantly less harassment under the second gender discrimination item (GD2: “Someone in your job makes crude and sexual jokes, stories, or comments that describe women negatively”) compared to Michoacán men (p = 0.02). California women farmworkers experienced significantly more harassment under the first unwanted sexual attention item (USA1: “Someone at your workplace tries to talk to other workers about their sex life or sexual preferences, making those people uncomfortable”) than men (p = 0.01).

Women were more likely than men to report WSH in California (10 of 13 behavior-based WSH survey items). In contrast, women farmworkers in Mexico were more likely than men to report WSH for only four of 13 behavior-based WSH survey items. WSH experience was greater in Mexico compared to California for men (all behavior-based WSH survey items) and for women (11 of 13 behavior-based WSH survey items).

Frequencies of Sexual Harassment Experiences with Behavior-Based Survey Items

Although many behavior-based WSH survey items never occurred, farmworkers also reported these items occurring as frequently as always (1 to 2 times a week), very frequently (1 to 2 times a month), frequently (4 to 6 times a year), and sometimes (1 to 3 times a year) (fig. 1). A type of WSH that was always (1 to 2 times a week) experienced was gender-discrimination among a few men and women in California and Michoacán. Another WSH type that always occurred was quid pro quo for a few men in Michoacán. The types of WSH that occurred very frequently (1 to 2 times a month) were gender discrimination, hostile behavior, and unwanted sexual attention among a few men and women in California and Michoacán. Another WSH type that occurred very frequently (1 to 2 times a month) was quid pro quo among women (2%) in California and women (2%) in Michoacán.

Figure 1.

WSH frequency of gender discrimination (GD), hostile behavior (HB), unwanted sexual attention (USA), and quid pro quo (QPQ) among women and men farmworkers in California (U.S.) and Michoacάn (Mexico). Asterisks (*) indicate that women are significantly different (p < 0.05) from the respective men based Kruskal-Wallis testing. Levels are never (0 times), sometimes (1 to 3 times a year), frequently (4 to 6 times a year), very frequently (1 to 2 times a month), and always (1 to 2 times a week).

The types of WSH that occurred frequently (4 to 6 times a year) were gender discrimination, hostile behavior, and unwanted sexual attention among men and women in California and Michoacán. Only quid pro quo occurred frequently among a few men and women in Michoacán and women in California. The types of WSH that occurred sometimes (1 to 3 times a year) were gender discrimination, hostile behavior, unwanted sexual attention, and quid pro quo among men and women in California and among men and women in Michoacán. We did not detect statistically significant differences in WSH frequency by gender within a country in our Kruskal-Wallis testing.

Perpetrators of Sexual Harassment Experiences with Behavior-Based Survey Items

Many farmworkers did not indicate the work role or position (e.g., coworker, supervisor) of perpetrators (fig. 2). These missing values were greatest for quid pro quo in Michoacán. Despite these missing data, the findings revealed that WSH was perpetrated by both coworkers and leadership. Coworkers were more frequently responsible for gender discrimination, hostile behavior, and unwanted sexual attention than leaders among all subgroups. Leaders were more frequent perpetrators of quid pro quo than coworkers.

Figure 2.

Perpetrators of all types of workplace sexual harassment (WSH) experiences, including (a) gender discrimination (GD), (b) hostile behavior (HB), (c) unwanted sexual attention (USA), and (d) quid pro quo (QPQ) among women and men farmworkers in California (U.S.) and Michoacán (Mexico).

Mixed-Gender Sexual Harassment Experiences among All Farmworkers

Combinations of perpetrator and target included men targeting women, women targeting women, men targeting men, and women targeting men (fig. 3). Data revealed greater men targeting women than women targeting men. Twenty-one percent of all men in California, 41% of all women in California, 35% of all men in Michoacán, and 34% of all women in Michoacán reported experiencing or witnessing men targeting women.

Figure 3.

Mixed-gender combinations of perpetration of workplace sexual harassment (WSH) among 197 farmworkers in California (U.S.) and Michoacán (Mexico).

Discussion

We surveyed farmworkers in a cross-sectional study to investigate workplace sexual harassment (WSH) among 197 women and men farmworkers in California (U.S.) and Michoacán (Mexico). Demographics, occupational characteristics, and workplace infrastructure findings offer valuable information. The workers in California were slightly older, had more education, were more likely to be paid by the hour, and reported slightly higher availability of supervisor help than the workers in Michoacán.

WSH experiences revealed important differences and similarities between men and women farmworkers in California and Michoacán. For example, more California victims faced retaliation (97% to 100%) than victims in Michoacán (50% to 60%). However, behavior-based questioning revealed slightly higher levels of quid pro quo type WSH in Michoacán (5% to 12%) than in California (3% to 7%). Direct inquiry-based (“Have you ever been the victim of or bystander to workplace sexual harassment?”) WSH among women (49%) and men (21%) in California and among women (7%) and men (13%) in Michoacán and behavior-based WSH accounts (using explicit examples of WSH behaviors perpetrated against the participant or witnessed by the participant as bystander) among women (as high as 53%) and men (as high as 45%) in California and among women (as high as 65%) and men (as high as 68%) in Michoacán document WSH exposure in the previous year. The frequency of WSH experiences and the perpetrators in the behavior-based survey items revealed that WSH is a common problem involving coworkers and leadership positions.

The demographics of the California participants were similar to the U.S. National Agricultural Workers Survey (Hernandez and Gabbard, 2019). Farmworkers in Mexico had comparatively lower levels of education, were younger, and spoke Purhépecha, an indigenous language. Most farmworkers were married, had children, and worked among family. This civil status and work crew composition may influence workers to maintain group harmony. These dynamics can be for good (encouraging workers to look out for one another, defending each other when one is attacked, and supporting each other when they want to report WSH) and ill (keeping victims silent because the harasser is a family member or because of the fear that reporting may lead to the entire family facing retaliation through firing, social stigma, or public shaming).

Farmworkers spoke English, Spanish, and Purhépecha. Studies have shown that indigenous women farmworkers are often unable to access or understand WSH prevention training because the trainings are not available in their language (Murphy et al., 2015). Our findings support efforts to partner with indigenous community groups and advocates to help translate training tools into audiovisual forms so that workers can understand and access materials on WSH prevention.

California participants reported being the target of or witness to greater WSH than Michoacán participants based on direct inquiry. In addition, gender differences were more pronounced among California farm workers. This could be due to California farmworkers’ greater exposure to WSH training and greater assimilation in comparison to Michoacán workers, which allowed them to identify WSH and label it. Michoacán men also more frequently reported being the target of or witness to WSH than Michoacán women based on direct inquiry. The opposite was seen in California. More California women reported being the target of or witness to WSH than California men based on direct inquiry.

The reason for women reporting greater WSH than men in direct inquiry in the U.S. may be because women who migrate end up participating in positions beyond domestic roles, such as in the workforce, where they receive workplace training on their rights (Cervantes-Pacheco et al., 2011). Greater participation in the workforce offers more legal protections against sexual harassment, which may explain the higher levels of reported WSH experiences among California women compared to California men and higher levels of reported WSH experiences among Michoacán women compared to Michoacán men based on direct inquiry. Lastly, direct inquiry revealed that Michoacán victims were forced to change their jobs less than California workers. This may be due to Michoacán workers’ reduced access to alternative positions and lower pay in comparison to California workers.

Michoacán farmworkers reported being the target of or witness to WSH more frequently than California participants in the behavior-based survey items. This could be due to how the items were structured. For example, in the behavior-based survey items, the Farmworkers were not required to label WSH incidents as such; rather, they reported if a behavior occurred that validated surveys have classified as WSH. It is important to ask behavior-based survey items because some WSH victims do not label their experiences as WSH, and their experiences therefore go undetected (Shupe, 2019). Our findings from the behavior-based survey items may reflect a true higher occurrence of WSH among farmworkers in Michoacán compared to farmworkers in California. Both direct-inquiry and behavior-based survey techniques have been used to measure WSH prevalence here and in other studies because behavior-based surveys pose some concerns about the generalizability of findings; thus, it is important to include both direct-inquiry and behavior-based measures of WSH (Willness et al., 2007).

Only half of farmworkers reported WSH in direct-inquiry questioning, and all but one perpetrator went unpunished. Farmworkers have stated that operators must improve on enforcing policies and ensuring consequences (Prado et al., 2020). Researchers have also argued that sexual harassment should be addressed by the organization as a whole and include prevention, rather than simply focusing on improving employees’ skills for dealing with harassment (Fitzgerald and Shullman, 1993). These findings suggest that some agricultural organizations must address how to follow through with ensuring consequences for offenders from the policy and leadership levels.

Behavior-based WSH experiences differed between countries. Michoacán participants reported greater exposure to WSH than California participants. Gender discrimination, unwanted sexual attention, and hostile behavior were more prevalent than quid pro quo in both California and Michoacán. This is similar to other studies in which all four WSH types were also reported (Kim et al., 2016; Murphy et al., 2015; Waugh, 2010). Stratifying by gender revealed that men and women in California and Michoacán experienced WSH. Both men and women reported WSH experiences as target or witness/bystander. Recent research has shown that WSH can negatively impact men and women farmworkers equally (Prado et al., 2020).

Previous surveys on WSH among women farmworkers showed similarities with our sample and suggested that WSH is not solely a sexual problem springing from natural feelings of sexual attraction. Among a group of previously surveyed California farmworker women, researchers found that the most frequently reported type of WSH was gender harassment or gender discrimination, second was unwanted sexual attention, and last was sexual coercion or quid pro quo (Waugh, 2010). Our subgroup of California women farmworkers was similar because they also reported the highest WSH exposure for gender discrimination, second for unwanted sexual attention, and least for quid pro quo. Therefore, the most common kind of WSH among California farmworker women was gender discrimination, which is conduct aiming to establish dominance, not pull victims into sexual activity (Cortina and Areguin, 2020).

Victims reported that they faced WSH annually, monthly, weekly, and even daily. Michoacán and California farmworkers experienced similar frequencies of gender discrimination and hostile behavior. California farmworkers reported more frequent unwanted sexual attention than Michoacán farmworkers. This difference may be due to how Farmworkers in the U.S. receive training that can make them more informed about WSH and thus more likely to report it (Murphy et al., 2015). Michoacán farmworkers reported more frequent quid pro quo WSH than California farmworkers. These findings may be due to the lack of human rights awareness, lack of WSH education, increased worker exploitation, and increased vulnerability in Mexico (Andrade-Rubio, 2016).

Farmworkers reported exposure to all gender combinations of WSH perpetration in which both men and women were harassers and victims. The mixed-gender combinations of men targeting women and women targeting women were the most frequent combinations reported. These findings offer evidence that men can be victims and women can be perpetrators. However, these findings also confirm prior research showing that farmworker women are disproportionately victimized and that men are the major perpetrators (Andrade-Rubio, 2016; Arellano Gálvez, 2014; Kim et al., 2016; Murphy et al., 2015; Waugh, 2010). Our previous work with farmworker focus groups used vignettes illustrating mixed-gender combinations to promote discussion about discrimination and harassment against lesbian and gay workers in agriculture (Prado et al., 2020; Villegas, 2019). Further research is needed to examine the kinds of WSH that LGBTQ (lesbian, gay, bisexual, transgender, and queer) workers face and their unique vulnerabilities.

We confirmed that WSH perpetrators included coworkers and leaders (Waugh, 2010). More coworkers than leaders were identified in perpetrating gender discrimination, hostile behavior, and unwanted sexual attention. Some of the men and women farmworkers did not identify the work role or position of some perpetrators. This may reflect not knowing the perpetrator’s position or fear of retaliation. None of the Michoacán participants wanted to indicate the work role or position of the perpetrators of their quid pro quo experiences. Unwillingness to provide information about perpetrators among women farmworkers in Michoacán may be due to reduced occupational, social, and cultural support. For example, women may face severe WSH in the workplace and in the home and receive blame for the WSH perpetrated against them. Women are often expected to endure the oppression of men because of the assumption that men have higher authority and are superior (Andrade-Rubio, 2016; Arellano Gálvez, 2014; Calvario Parra, 2016; Zúñiga-Elizalde, 2008).

Conclusions

Women and men farmworkers in Mexico and the U.S. were exposed to multiple kinds of WSH. These experiences occurred frequently. Perpetrators were found at the coworker and leadership level. Few perpetrators were punished, whereas victims often reported retaliation and having to change their jobs.

Farmworkers are essential, yet vulnerable to exploitation in the agricultural industry. This study lifts their voices on the violence perpetrated against them. Although women farmworkers are the fastest-growing group in agricultural labor (Hernandez and Gabbard, 2019), they are the primary victims of sexual harassment and are multiply marginalized based on race, class, and sex (Kim et al., 2016; Murphy et al., 2015; Villegas, 2019; Waugh, 2010). Agricultural crews often include many members of the same community or family (Waugh, 2010). Both men and women in families can be directly affected as victims or as bystanders (Bauer and Ramirez, 2010). Lastly, retaliation in response to reporting can include the firing of not just the victim but also the victim’s family, further silencing victims (Bauer and Ramirez, 2010).

Future research should introduce new and expanded ways of thinking about this problem. Other studies could contribute information on the psychological impacts of victims’ experiences and help clarify how men and women differ as WSH victims and as bystanders. This supports efforts to garner more participation from men in roles other than as perpetrators (Prado et al., 2020). Other studies could contribute information on employer or leadership experiences, perspectives, and beliefs regarding WSH. This would provide insight from higher levels of the organization, including those with responsibility for working conditions. Employers may face challenges or social barriers to communicating policy, reprimanding offenders, and assisting victims. Understanding of experiences and attitudes at the leadership level is critical because policy is made, executed, and monitored at this level. Lastly, other studies could contribute information on the WSH experiences of LGBTQ+ farmworkers. The discrimination that members of this group face has yet to be fully evaluated or understood.

Strengths of this study include the use of our community advisory boards (CABs), inclusion of both men and women farmworkers, our binational comparison, and our WSH measurement approach. The CABs consisted of farmworker women, farmworker welfare advocates, non-profit representatives, legal entities, industry officers, and community leaders. The CABs contributed to the study strategies, materials, and dissemination. To our knowledge, this study is the second to use surveys in the evaluation of WSH in agriculture and the first to include both men and women farmworkers. We provide quantitative data that supports earlier qualitative findings (Andrade-Rubio, 2016; Kim et al., 2016; Murphy et al., 2015). In particular, our comparison of California and Michoacán offers a binational view. Evaluating findings in two countries facilitates comparisons across cultures. This transcultural analysis broadens the impact of the findings on prevention efforts in Mexico and the U.S. (Valentín et al., 2005). These comparisons are relevant because Mexico is the country that most frequently sends farmworkers to California and other regions within the U.S. Lastly, we used both direct-inquiry and behavior-based survey items to measure WSH, an approach that is considered appropriate in collecting the full range of participant WSH experiences (Willness et al., 2007).

Limitations of this study include potential biases from our sampling method, participant self-selection, and limited geographic scope. Our sampling method consisted of in-person purposeful and convenience recruitment, which could have led to an overrepresentation of a specific response type. Although our sample was similar to the general population of farmworkers in national and regional surveys (Hernandez and Gabbard, 2019), small demographic differences, such as older men in California and more unmarried women in Michoacán, may have introduced variation into our sample. Limited funds prevented large-scale sampling and restricted our geographic scope. However, there was notable diversity in the representation of experiences because we included (1) Spanish and Purhépecha speakers, (2) men and women, and (3) residents of Mexico and the U.S. Additionally, some statistically significant findings may represent Type I errors because we did not correct for multiple comparisons.

Overall, the results of this study help to raise awareness, facilitate development of educational materials, and suggest policy directions for the prevention of WSH in agriculture.

Highlights.

• Women and men farmworkers reported workplace sexual harassment (WSH).

• WSH occurred as frequently as daily.

• Both coworkers and leadership were perpetrators of WSH.