Brief background of current measurement of harassment climate

Extant measures of harassment climate have been based on Hulin, Fitzgerald, and Drasgow’s (Reference Hulin, Fitzgerald, Drasgow and Stockdale1996) pioneering work in this area. Their measure and subsequent formulations of it are collectively referred to as Organizational Tolerance for Sexual Harassment (OTSH). Scholars who have relied upon the OTSH have reported consistently high scale reliabilities, with alphas ranging from .88 to .95 (c.f., Bergman, Langout, Palmieri, Cortina, & Fitzgerald, Reference Bergman, Langout, Palmieri, Cortina and Fitzgerald2002; Goldberg, Rawski, & Perry, in press; Kath, Swody, Magley, Bunk, & Gallus, Reference Kath, Swody, Magley, Bunk and Gallus2009; Williams, Fitzgerald, & Drasgow, Reference Williams, Fitzgerald and Drasgow1999). Despite the soundness of the item content and the psychometric properties of the scale, a closer look suggests that the scale may not fully capture the knowledge we have gained, as a field, in the 20 years since it was first developed.

Gaps and solutions to improve the measurement of sexual harassment climate

In this section, we call for researchers to consider whether the shared-perception view of OTSH is the most meaningful analytical approach to use. Although scholars have used different variants of OTSH, the scale typically comprises seven to nine items that assess employees’ perceptions of the extent to which organizations (a) pursue preventative actions, (b) thoroughly investigate complaints, (c) enforce penalties against harassers, and, more generally, (d) allow harassers to get away with their behavior. We identify the limitations in the measurement of each of these subcomponents and offer insights into its improvement.

OTSH is considered to be a climate measure (i.e., shared perception). How a person perceives a behavior matters more than the intent of that behavior, and if everyone agrees that the organization does not take sexual harassment seriously, then there is likely little concern about whether the perception is accurate. However, in many organizations, harassment is something of a well-known secret, where such behaviors are seen as harmless, “locker room talk,” to everyone but the victims. As such, we need to question whether looking at aggregate perceptions is the most useful means of understanding the situational causes and outcomes of sexual harassment. It may be the case that the outlying observation of the victim is more telling than the normative observations of those who give tacit approval to the harasser by excusing his/her behavior.

The high reliability of the OTSH is somewhat surprising, given that it captures multiple facets of handling sexual harassment complaints. However, as state legislatures are increasingly forcing organizations to provide training, without concurrent requirements for investigations and penalties, over time responses to these items are apt to diverge. That is, respondents may indicate that the organization has engaged in preventative action (because it was legally required to do so) but still believe that the organization is doing little to correct harassment (because it was not required to legally). As a field, we can, and should, prepare ourselves for this inevitable decrease in interitem reliability.

Another concern is that the prevention items in the current measure do not adequately capture the effectiveness of the organization’s actions. Medeiros and Griffith (Reference Medeiros and Griffith2019) aptly note that there are vast differences in how organizations conduct sexual harassment training and that these differences (length, modality [online versus face-to-face], participant interaction) have substantial effects on trainees’ post-training knowledge, attitudes, and behavioral intentions. Yet, the only OTSH item that ostensibly captures whether organizations take training seriously or use it as a preventative measure is, “Actions are taken at my workplace to prevent sexual harassment.” An employee who does not hold a degree in industrial and organizational (I-O) psychology/human resources (HR) might reasonably believe that because he watched a half-hour video about harassment after joining the organization that the organization does take actions to prevent harassment even though research suggests that this action is not apt to have any effect. Indeed, even among employment attorneys, it is common practice to rely upon the existence of any training program as a basis for a Faragher v. City of Boca Raton and Burlington Industries, Inc. v. Ellerth affirmative defense, even though most scholars would balk at the notion that a 30-minute video shown upon hire constitutes a “reasonable step” to prevent sexual harassment.

In a similar vein, the OTSH has but one itemFootnote ¹ devoted to investigations, “My workplace provides thorough investigations of sexual harassment complaints.” One limitation of this item is that to answer it accurately, one would be required to have knowledge and some experience with the investigation process. Given that most victims of sexual harassment do not report, few respondents would have even the most basic insights upon which to base such an assessment. In the first author’s experience as an expert witness, it is often the case that the stated procedure for investigating harassment complaints within an organization will bear little resemblance to the actual investigation process that is used in cases that subsequently result in litigation. For example, an organization’s policy may state, “We thoroughly investigate all complaints”; however, the investigative file suggests that there were no witnesses questioned. Thus, relying on an organization’s stated policies as the basis for determining whether one’s workplace provides a thorough investigation is a weak proxy for having participated in the process.

Further, even for employees who have participated in an investigation (whether as a complainant, respondent, or witness), on what basis do they determine whether it was thorough? The recent visibility of Justice Kavanaugh’s hearings make clear that the bar for conducting a “thorough investigation” varied from senator to senator. The Equal Employment Opportunity Commission’s (EEOC) Enforcement Guidance (1999) sets forth general criteria: that investigations need to be prompt, thorough, and impartial. It also provides questions to ask the complainant, the alleged harasser, and third parties. These outlined procedures emphasize the importance of making credibility assessments, having guidance for making the assessments (including cases involving allegations for which there are no eye witnesses, as is often the case for sexual harassment), making a determination, and taking corrective action. Yet, although the EEOC’s guidance has prompted a plethora of practitioner articles on conducting sexual harassment investigations, with one recent exception (Goldberg et al., in press), there has been no scholarly research addressing the topic. In their empirical study, Goldberg and colleagues found that in order for HR managers to change their attitudes and acquire appropriate knowledge about effectively conducting sexual harassment investigations through training, organizations need to foster a climate that is intolerant of sexual harassment.

The OTSH also measures employees’ perceptions of organizations enforcing penalties against harassers and allowing harassers to get away with their behavior. However, knowledge of this would require that harassment outcomes (and by extension, claims) be publicized, which is often not the case. In a comprehensive study that tracked more than 1,600 employment discrimination cases, including sexual harassment cases, it was found that a very small percentage actually makes it to trial (Nelson, Nielsen, & Lancaster, Reference Nelson, Nielsen and Lancaster2007). Organizations are often motivated to handle cases outside of court to avoid bad publicity. Alternative Dispute Resolution (ADR) is used as a mechanism in which disputes can be settled outside of the official legal system (Stockdale & Sagrestano, Reference Stockdale and Sagrestano2010). However, one notable drawback is that because disputes are settled confidentially, no one except the immediate parties is aware of the outcome (Beiner & O’Connor, Reference Beiner, O’Connor, Crosby, Stockdale and Ropp2007). In order for the OTSH measure to accurately capture perceptions of organizations’ penalties toward harassers, decisions need to be widely known. The practicality of such is challenging, and at the very least, the complainant has a right to know if actions were taken pursuant to his/her complaint.

Finally, given the current societal effect of the #MeToo movement, it is clear that broader contextual factors cannot be ignored. Indeed, because of the momentum of this movement, EEOC has reported a significant increase in workplace sexual harassment claims. Consequently, this shift is likely to influence more global perceptions measures in the OTSH as well. We suggest that expanding these items to adequately capture this change would lead to a separate subfactor, distinct from prevention (i.e., training) and response (investigations and corrective actions), but that the three would be correlated with one another and all would be important to better understand an organization’s tolerance for sexual harassment.