The Safest Bet: Identifying and Assessing Risk in Faculty Selection

Introduction

Imagine members of a faculty search committee convene to discuss candidates. The chair, per university policy, reads from a recently approved equity charge, reminding the committee that their deliberations should not discriminate based on gender, race, or other protected categories. The chair then opens a discussion about their search. A committee member raises a concern about timing, asking the chair if they think the committee will be able to set up the first round of interviews in the next month. The committee member is worried because they know several “hot” candidates are already interviewing at other institutions—their search might lose out. The chair says yes, although they are more concerned about finding candidates with expertise in a [particular research method]. The committee crafted the job description to cast a wide net of expertise, an inclusive hiring strategy suggested by their associate dean. Yet, the only faculty member who uses [the particular research method] just retired. The chair hopes they could select someone with similar training because the department rarely gets a new search. Two committee members raise separate concerns about the rubric by which the committee will evaluate candidates. One suggests that the rubric is not sensitive to subfield differences, noting that Subfield A prefers candidates who have completed postdocs while other subfields do not, which shapes number of publications and research funding. The second committee member, a graduate student, observes that the rubric does not emphasize teaching, but wonders how important that is given their institution’s research emphasis. The chair sighs, acknowledging that “the process is not perfect, we just need to do the best we can with it and move forward.”

This scene is familiar to many. Faculty hiring is a complex cultural practice that occurs with individual, institutional, disciplinary, and societal forces, shaping it simultaneously (Posselt et al., 2020). The vignette depicts a search for a faculty member in a particular discipline, and the norms of that discipline shape how committee members view methodological expertise, postdoctoral experience, and publications in distinctive ways (Posselt, 2016). At the same time, the search occurred at a research-intensive university, which shapes expectations for research qualifications compared to teaching (Gonzales & Terosky, 2016). The “wider net” job description, equity charge at the beginning of the meeting, and use of rubric occurred as part of institutionally mandated inclusive hiring policies widely adopted within higher education (Stewart & Valian, 2018). Universities are also located in states with different higher education funding and social policies that shape how many faculty members can be hired and who is interested in applying (Stewart & Valian, 2018). Finally, each search committee member brings individual tastes and preferences (White-Lewis, 2020). The student member perhaps cared more about teaching because this faculty member might be their instructor. The chair wants someone in a certain area and is worried that they may not be able to replace their colleague. As human beings, search committee members also brought cognitive and social biases (Eaton et al., 2019). Social biases might not have explicitly appeared in the questions committee members asked, but they would inevitably be part of how committee members evaluated candidates, considered their merit in light of qualifications, and whether they thought the candidates “fit” the job posted (Eaton et al., 2019; O’Meara et al., 2020). In sum, the criteria, processes, and outcomes of faculty evaluation will be shaped by each of these contexts (Possselt et al., 2020).

In this study, we argue that how faculty search committee members identify and assess risk is an overlooked yet critically important context that shapes faculty hiring outcomes. Broadly speaking, risk assessment and identification refers to the process by which a decision-maker considers a future activity and, faced with uncertainty, perceives that something negative could happen. Decision-makers are often irrationally more sensitive to and influenced by the perceived negative outcome because they fear loss (Kahneman, 2011). In the vignette above, the committee raised two risks: the risk that they would not move fast enough and thus lose the best candidates, and the chair’s perceived risk that they would not find someone from a particular subfield. In both cases, the committee honed in on the negative alternatives and the potential loss. This reflects research that shows that when search committee members predict something “bad” could happen if they choose a particular candidate (e.g., the candidate says no because the search moved too slowly and the search fails), they are less likely to choose them (Rivera, 2017).

However, the identification and assessment of risks is not always neutral. Decision-makers bring cognitive and social biases to the process of identifying and assessing risk in ways that implicitly advantage men and white candidates over women and racially minoritized candidates (Posselt et al., 2020). For example, if the subfield that the chair prefers is also a subfield with fewer minoritized candidates, that narrowing may reduce the diversity of those considered “meritorious candidates” (White-Lewis, 2021). If the committee is worried about losing out on the “hot candidates,” they may rush the process and overlook a candidate who meets the specific departmental needs. Because risk identification and assessment are so engrained in faculty evaluation repertoires of practice—or the array of activities commonly associated with being an academic—they have largely escaped scrutiny (Gutierrez and Rogoff, 2003, as cited in Posselt et al., 2020). By pulling risk into view and considering the context(s) that shape it, academe and faculty members can transform evaluation practice(s) and potentially mitigate the inequities that have long plagued the professoriate (Posselt et al., 2020).

The purpose of this study was to understand how committee members identified and assessed risk in faculty selection. We were especially interested in revealing the kinds of risks identified, and how committee members assessed them during faculty searches. Drawing from the work of Posselt and colleagues (2020), who articulated a framework for equitable evaluation and decision-making in higher education, we approached this study from both a constructivist (e.g., assuming search committee members were constructing risks, individually and collectively, often unconsciously, and without realizing how they shaped decisions) and critical and power-analytic perspective (e.g., assuming that the identification and assessment of risks can reinforce power asymmetries laden with social and cognitive biases that are not race and gender neutral).

Our research makes distinct contributions to the literature by (a) adding to a very small number of studies that have examined search processes from the inside using ethnographic observations, (b) identifying particular kinds of risks and how those risks were assessed in five faculty searches and (c) scrutinizing how the identification and assessment of risk are not neutral but shaped by multiple contexts, including structural and individual biases, and have the potential to reproduce inequities. These contributions are important because it is only by getting inside faculty searches and seeing the criteria and mechanisms shaping actual decisions that practical recommendations for more effective and inclusive searches can be rendered (Liera & Hernandez, 2021). We argue that perceptions of risk are tied to structural biases that evade discussions of implicit bias. That is, although overt discrimination does occur, more often structural biases are carried forward as part of risk assessments that are viewed as objective or neutral (Ray, 2019). By revealing how committees identify and assess risk in faculty searches, and by revealing potential ways in which these processes hold “risk” of disenfranchising minoritized groups, we can make recommendations for helping search committees exhume perceived risks, and therefore enhance equity, before searches begin.

In the next section we review the literature on faculty evaluation and selection. We situate our study in prior work showing how faculty hiring decisions are created within cultural systems of practice, and thus shaped by power asymmetries, disciplinary preferences, and cognitive and social biases that disadvantage minoritized groups. We then turn to social and behavioral science research on risk identification and aversion and consider how biases might be “carried in” by these innate and seemingly—but not—neutral processes in faculty selection.

Literature Review: Key Contexts Shaping Faculty Evaluation and Selection

In reviewing the extant literature on faculty selection, we were influenced by the work of Posselt and colleagues (2020), who identified a framework for equitable evaluation and decision-making in higher education. Most faculty searches engage in evaluation through formal systems and ad hoc judgements—two forms of evaluation that shape each other and can advance equitable outcomes and/or reproduce inequities (Posselt et al., 2020). Research has shown how preferences emerge in both spaces and can advantage certain candidates and disadvantage others (White-Lewis, 2020). Formal evaluation criteria that weigh publications more heavily may advantage white and male candidates (Posselt et al., 2020), while a committee may decide on an ad hoc basis that they prefer someone with a subfield specialty not previously articulated (White-Lewis, 2020).

We use three key points made by Posselt and colleagues (2020) to outline research on faculty selection most relevant to our study. The first point is that evaluation is a process in which evaluators use “shared interpretive schemas and scripts” that “sort people, scholarship, or academic institutions” into hierarchical, value-laden categories (Posselt et al., 2020, p. 10). In faculty searches, evaluators seek to stratify and rationalize qualifications through the seemingly objective process of differentiating between candidates (Lamont et al., 2014). This sorting happens within a cultural community influenced by multiple factors, as illustrated in the opening vignette. Search committees identify evaluation criteria, engage in deliberative processes, and determine outcomes embedded in specific cultural communities that have developed “repertoires of practice” (Gutiérrez and Rogoff, 2003, cited in Posselt et al., 2020) over time. The repertoires search committees use, such as having the chair organize the meeting, become so engrained as processes and, later, grounds for judgements, that they are rarely questioned (Lamont et al., 2014). However, repertoires of practice are nested in individual identity, networks, society, disciplines, and institutions (O’Meara et al., 2020; Posselt et al., 2020; White-Lewis, 2020). For example, search committees rely heavily on trust networks and information from known colleagues when evaluating academic leaders or potential doctoral students (Posselt, 2018). This tendency to advantage candidates who have worked with prestigious mentors, or those candidates who are well known in predominantly white institutions, then becomes viewed as the “natural order” that disadvantages candidates from Minority Serving Institutions (Posselt, 2018).

The second point situates faculty selection as socially constructed and rooted in contextual, sociocultural forces. Search committee members will rarely be influenced by a single priority but rather a “hierarchy of priorities and preferences” will drive evaluation criteria, processes, and outcomes (Boltanski et al., 2006, p. 21). Often, disciplinary and institutional forces intersect to shape faculty selection. A mission-driven, religious institution’s cultural emphasis on niceness, for instance, shapes interactions between search committee and candidate and can undermine efforts at making faculty hiring more racially equitable (Liera, 2020). Socialization in research universities produces notions of legitimate scholarship that discount applied or interdisciplinary scholarship (Gonzales & Rincones, 2012) and, thus, search committees may not advance candidates who do this type of work. Similarly, judgements of “talent” or “brilliance,” rarely listed in formal job criteria, are often informal expectations for faculty positions (Lamont, 2009). Given that faculty members associate talent or brilliance with those who attended elite institutions (Bell & Chong, 2010)—institutions to which minoritized groups have less access (Freeman & DiRamio, 2016; Posselt, 2018)—we expect there to be social and cultural aspects that are both difficult to disentangle but that nonetheless concurrently influence faculty hiring.

The third point is that structural and individual bias is built into faculty evaluation and the processes and conditions surrounding it (Posselt et al., 2020). For example, Black women applying for administrative positions in universities are subject to more rigorous “filters” or sets of criteria compared to white women and men (Danowitz Sagaria, 2002). Departmental “silos” may hinder the hiring of scholars who do work across disciplines (Gonzales & Rincones, 2012). Gendered perceptions of women in heterosexual relationships shape assessments of candidates for faculty positions (Culpepper, 2021; Rivera, 2017). In other words, structural aspects of the faculty work environment as well as individual predispositions can be biased and therefore produce unequal evaluation results (Posselt et al., 2020).

Theoretical Framework: Perceptions of Risk and Loss Aversion

Theories of how decision-makers identify and assess risk (Slovic, 2000), and how they make decisions to avoid risk (Kahneman, 2011) from the fields of behavioral economics, social and organizational psychology, and decision-making, guided this study. In any given decision, individuals try to maximize gain (or “win”) and minimize loss, thereby advancing their self-interest and managing or averting risk (Kern & Chugh, 2009; Samuels & Zechhauser, 1988). However, often we are not assured that the path we choose (the “alternative”) will result in a win or loss. Prospect theory illustrates how individuals make decisions under such uncertainty. It states that we assess alternatives relative to reference points, or the status quo (Kahneman, 2011). Alternatives that deviate from our reference points are often considered riskier, especially if they could result in loss (Kahneman, 2011). Moreover, we tend to weigh prospective losses more heavily than equivalent gains (Kern & Chugh, 2009). We will choose the status quo, foregoing any gains, to minimize loss (Kern & Chugh, 2009; Samuels & Zechhauser, 1988).

In this study, we consider the nested contexts noted above to be the reference points that inform search committee members’ views on what is considered risky. Faculty searches develop formal criteria to rate candidates and identify a shortlist of candidates with the greatest number of characteristics they desire—a strategy to maximize gains and reduce risk. Reference points inform the creation and weighting of these criteria and the ad hoc assessments that happen as evaluators apply them. Thus, even if a certain candidate meets or exceeds most criteria, the committee might consider a candidate a “loss” if they have not met a specific criterion (e.g., received an award), especially if other faculty members in their department did.

Social biases, widely observed in faculty hiring (O’Meara et al., 2020; White-Lewis, 2020) also inform reference points. Experimental studies show that, across different academic appointment types, faculty members view candidates whose names are perceived to be men and white (and sometimes Asian) to be more competent and hirable compared to candidates perceived to be women, Black, and Latinx. (Beattie et al., 2013; Eaton et al., 2019; Moss-Racusin et al., 2012; Steinpreis et al., 1999). Additionally, qualitative studies wherein researchers observe faculty search committees show gendered assessments of “moveability” (Rivera, 2017) and racialized conceptions of “fit” (White-Lewis, 2020) that undermine the hiring of women and racially minoritized candidates. Said another way, a candidate’s identity itself can become a reference point, becoming the anchor around which evaluations are made.

In theory, faculty searches are about adding to the department—a gain. However, much about the context surrounding hiring makes the process risk laden. Departments covet new tenure track faculty positions as a key resource, especially since they are not awarded easily (White-Lewis, 2021). Most tenure-track faculty searches are competitive, with anywhere between 40 and over 200 applications for a single position. Candidates each have different strengths and weaknesses, and job qualifications often leave room for discretionary judgment. As such, committees develop simple strategies to eliminate choices and/or sort the candidates: assessing risks is one rational way to make decisions in a complex environment (Lipsky, 2010).

However, such decision-making becomes problematic when we consider the connection between perceived risk, loss aversion, and the structural and individual biases that shape hiring decisions. Faculty search committee members might have a stated desire to increase faculty diversity, but then view candidates through a lens where perceived social identity is closely associated with perceived risk. For example, Rivera (2017) found that search committee members assessed women faculty candidates with partners to be “riskier” based on the long-standing gender stereotype that women with men partners often reject job offers. The perceived risk was that the search would fail; as such, the committees reduced risk by choosing men candidates with similarly excellent qualifications and for whom the stereotype did not apply. This assessment of risk occurred informally (i.e., partner status was not a formal criteria) and seemed rational based on the committees’ desire to reduce loss and maximize gains. Similarly, Settles and colleagues (2018) found that faculty of color reported that their research methods, topics of study (e.g., marginalized populations), and/or standpoints (e.g., social justice), are devalued in faculty promotion and tenure. As such, committees may view candidates who do this kind of work, most likely to be minoritized scholars, to be at greater risk for not making tenure or promotion. This kind of assessment then makes minoritized candidates a double risk—through the focus of their scholarship, a structural bias, and through their identity, an individual and societal bias.

Faculty members have much discretion in the assessment of risk, and the stakes are high. Our review of the theory and the literature shows that risk identification and assessment in faculty evaluation is (a) inevitable and innate to the process; (b) has rational and irrational components; and (c) is a cultural process shaped by multiple sets of contexts, reference points, and biases that decision-makers bring to the task (Posselt et al., 2020). As such, the equity concern is not that a committee uses assessment of risk as a way to sort and advance candidates. Rather, as Slovic (2000) observes, the equity concern is the “exercise of power” (p. 689) that emerges as risk assessments become intertwined with structural and individual biases and disciplinary logics that frame competence, hireability, merit, and legitimacy, and the process by which committees associate candidates of minoritized identities with more risks, and non-minoritized candidates with more gains. Our focus was to interrogate this murky territory of risk and its consequences for faculty hiring and equity. The research questions guiding this study were:

Methods

Site Selection and Case Descriptions

We gained access to five search committees within four different institutions via convenience sampling. Four committees were observed during the 2019–2020 academic year, and one committee was observed during the 2020–2021 academic year. The search committees were housed within departments that belonged to institutions participating in a National Science Foundation (NSF) grant focused on creating postdoctoral pathways to the professoriate. This increased upper administrative buy-in to access the searches; without this support it may not have been possible to observe the committees. However, the selected departments were not involved in particular interventions to increase inclusive hiring per se, although they were located at universities with institutional-level initiatives, which is typical. In this way, our five searches could be considered typical cases, rather than revelatory cases (Yin, 2018), which was intentional to ensure that the search processes we observed would be more theoretically “transferable” (Merriam & Tisdell, 2016) to other search processes at similar institutional types. Table 1 describes each committee; we do not provide the actual subfield of each due to confidentiality.

OPEN IN VIEWER

Table 1

Search Committee Descriptions

Engineering 1	Six-member committee in chemical engineering located at a research-intensive institution. The position description called for multiple tenured or tenure-track faculty across ranks, and emphasized a desire for DEI experiences. Their evaluation centered research with average consideration for teaching and above-average consideration of DEI competencies.
Engineering 2	Eight-member committee in environmental engineering located at a research-intensive institution. The position description called for a tenure-track assistant or associate professor relatively late in the academic job market. Their evaluation heavily centered research with minimal consideration for teaching and average consideration of DEI competencies.
Psychology	Four-member committee in developmental psychology located at a research-intensive institution. The position description called for a tenure-track assistant professor and associate professors. Their evaluation emphasized a balance of research teaching with above-average consideration of DEI competencies.
Biology 1	Five-member committee in biology located at a teaching-intensive institution. The position description called for an assistant professor who specialized in microbiology or related subfield. Their evaluation centered teaching and research, and service and DEI contributions to an extent.
Biology 2	Five-member committee in plant biology located at a teaching-intensive institution. The position description called for a tenure-track assistant professor. Their evaluation centered an above-average emphasis on research with standard emphasis on teaching and marginal consideration of DEI competencies.

Note. DEI = diversity, equity and inclusion.

Data Collection and Sample

We collected multiple sources of data to create a robust depiction of faculty search processes (Creswell & Poth, 2016; Merriam & Tisdell, 2016). This included ethnographic observations, document analysis, and interview data. First, to obtain administrative buy-in to access the searches, the first and third authors conducted informal interviews with search committee chairs. Through this we gained preliminary background knowledge on the department and search, outlined our researcher role to minimize concerns of search interference, and reviewed observation and consent procedures which were all approved by our university Institutional Review Board (IRB).

After obtaining approval, the third author attended search meetings in-person, by phone, and via video conferencing platforms (e.g., Zoom, Webex) and took notes using a protocol. Over the course of the observation the third author sat away from the committee and remained quiet so as not to interfere with search processes or serve as a reminder of their presence. Search committee members discussed a range of search-related topics over the course of the observations. This included generating and disseminating the ad, determining the selection criteria to apply to candidates, discussing the general pool of candidates, reviewing—and, at times, debating—candidate files in details to determine the short-list, Zoom interview performance, campus interview performance, finalist selection, and discussions of departmental votes. Four of the five searches occurred during the 2019–2020 academic year, and three of those secured their final hire prior to the onset of the COVID-19 pandemic. These meetings primarily happened in-person and occasionally over the phone or via Zoom. The fifth search occurred during the height of the pandemic, and they conducted their search exclusively over Zoom. All participants completed an informed consent form before being observed. Per our IRB agreement, the researcher did not write down the names of candidates, or have access to their application materials, which included their identities. Per institutional procedures, committee members were also not provided with candidate demographic information. However, committees often knew or inferred candidate social identity from prior knowledge of candidates, candidate self-reporting of this information in personal statements, and/or their involvement in particular identity-based affinity groups, and use of pronouns. Thus, we used committee member understanding of candidate social identities to form our own judgments for the purpose of analyses. ¹

Ethnographic observations were ideal to examine “the behavior, the language, and the interaction among members of [a] culture-sharing group” (Creswell & Poth, 2016, p. 90). The five searches reflected their distinct departmental histories, disciplinary norms, and institutional policies; they constituted separate “culture-sharing groups.” We used the same ethnographic observation protocol for each search. Table 2 provides the major domains of the ethnographic observation protocol with examples of risks identified and assessed throughout the searches. We drew inspiration from Posselt and colleagues’ (2020) domains of practice to create the protocol with a key interest in identifying where search committees believed risk to be and how it should be evaluated. This protocol guided the researcher through major domains of the faculty search, and served as the basis for later coding and analysis. For example, geographical context was raised as an issue (e.g., the perceived risk of a candidate not being able to do their research in a particular location) and in another search, the institution’s teaching focus meant there was a concern that a candidate would not completely “get” just how much time is spent on teaching once hired. These examples emphasize how institutional and state-level contexts inform perceptions of risk.

OPEN IN VIEWER

Table 2

Observation Protocol Domains With Risk Identification and Assessment Examples

Evaluation Criteria	Risk Identification: Disciplinary Expertise Risk Assessment From Biology 1: At multiple points throughout the search, it was a critical concern that the candidate has “the right tools,”“be a true [disciplinary expert]” and not “be on papers without doing the work.” At various times, committee members would use signals and cues from CVs and published work to determine if a candidate merely used software or made contributions to the study of that software or program.
Identities Discussed	Risk Identification: Interest Risk Assessment From Engineering 2: When discussing an international candidate before the COVID-19 pandemic was detected in the United States, one search committee member stated that they were a lower risk and would be easier to retain because “[their country] is a mess right now, they’ll definitely come.”
Short-List Formation	Risk Identification: Search Timing & Oversight Risk Assessment From Engineering 1: There were significant concerns early in the search that waiting until all applications were submitted before doing phone interviews would significantly slow down the process. In a later meeting the committee decided that they would switch to a rolling deadline so that they could conduct phone interviews prior to the application deadline.
Finalist Formation	Risk Identification: Interest Risk Assessment From Biology 1:“Interacting with her on an interpersonal basis, I didn’t get the sense she was interested in the job. She didn’t have any questions. When we asked questions, her responses were a couple of words. So not picking up on a desire to be here. Maybe she’s an introverted or shy person? But during an interview when you’re supposed to put on your best…”
Department & Institutional Considerations	Risk Identification: Collegiality Risk Assessment From Psychology: There was a concern that a finalist candidate was not interested in “understanding the department,” based on some of her comments that were taken as critiques of the department. “What would it look like to work with her if she’s going to jump to judgment without fully understanding?”

Note. CV = curriculum vitae.

In the overall study, we conducted 40 observations of search committee meetings, 18 observations of job talk seminars, and six observations of departmental meetings, for a total of 64 observations, to understand what individual and group-level practices facilitated and/or inhibited inclusive hiring efforts. Table 3 provides the data collection frequency, length of engagement, and aggregated search committee demographic information for the larger study. Additionally, we collected relevant documents, such as position descriptions, rubrics, and email chains between committee members, as appropriate. Finally, we conducted five interviews with each search chair at the end of their search, which also doubled as a form of member-checking. In these interviews, we recalled specific search details and confirmed who was hired after all departmental votes and related procedures were completed.

OPEN IN VIEWER

Table 3

Search Committee Observation Schedule and Demographics

	Engineering 1	Engineering 2	Psychology	Biology 1	Biology 2
General Meetings	9	5	12	3	11
Job Talks	3	2	6	3	3
Departmental Meetings	0	0	3	1	2
Duration of Engagement	7 months	6 months	10 months	7 months	7 months
Demographics Across All Five Committees*	Two Black men, one Latina woman, one Latino man, one South Asian woman, one South Asian man, eight white women, 13 white men.

Note. * Demographics aggregated across all five search committees to protect participant confidentiality.

Findings

All five of the searches identified risks related to candidate interest; candidate disciplinary expertise; candidate competence; candidate collegiality; and the timing and oversight of the search process itself. The first four risks emerged as committees made decisions about candidates, whereas the final risk existed throughout the search and shaped the overall process. Disciplinary, institutional, and individual contexts embedded within each search shaped risk assessment such that searches pulled different information and concerns forward, but all searches mulled over the five risks outlined in our themes. For example, all committees were worried that if the search did not move quickly, they would lose candidates to peer institutions. However, disciplinary norms shaped the time of year in which most searches were completed and whether the committee thought their search was going too slow or was ahead. Overall, our searches were more similar than different in what they identified as risks and how they assessed risk. However, as noted in our literature review, individual, organizational, and field norms all shape evaluation, and we acknowledge contexts each search uniquely pulled forward as relevant.

The organization of our findings is as follows. We have five risk subsections. In each of the five subsections we share (a) the nature of the particular risk—what the search committee was concerned might happen; (b) how the committee assessed the risk; and (c) whether and how individual, institutional, and disciplinary contexts and structural and individual biases shaped risk identification and assessment.

Discussion and Implications

“Although…dangers are real, there is no such thing as “real risk” or “objective risk” (Slovic and Weber, 2002, p. 4). Although embedded in and shaped by similar and distinct contexts, all five faculty search committees in this study were concerned with risk. They assessed whether candidates were truly interested in their position, whether the candidate had the pure expertise they wanted, whether candidates were competent enough, and whether they would be collegial colleagues. All searches worried about whether they were going fast enough and would lose the best candidates if they did not move faster, and whether an administrator would step in and take the decision away from them. In this section we expound on three points by relating these findings to our literature review and theories of risk aversion.

Our first point is that the identification and assessment of risks was integrated into formal and ad hoc assessments of qualifications and determinations of outcomes (Posselt et al., 2020) and was thus part of all five committees’“repertoires of practice” (Gutierrez & Rogoff, 2003). The process of identifying and assessing risks seemed in many ways, natural, necessary, and routine. Committee members did not question the desire to, for example, make sure candidates were really interested in their positions or would be good colleagues. At face value, these qualities seemed like reasonable things to expect of a new colleague and would stave off the losses the department would experience if the opposite were true. To some degree, the risks gave the appearance of being neutral, objective, and fair. With over 100 applications in some cases, this professional, discretionary task required some way to differentiate between candidates (Lipsky, 2010; Posselt et al, 2020). The assumption that speed would help the committee achieve their goal of obtaining the best candidates was only disrupted by the Psychology committee. While this committee identified that timing was critical, they decided to discuss each candidate anyway, an inclusive hiring practice often encouraged by institutions (Stewart & Valian, 2018).

At times, committee members’ reasons for selecting or deselecting candidates explicitly mentioned risks, and in other cases the risks seemed to be carried in with formal criteria for the position. For example, in one case, the committee learned that one of their candidates had received an offer from another institution. This caused discussion as to whether this candidate was likely to leave their search, and whether their competitiveness made them more likely to leave their position later. The committee spoke about this risk overtly as part of their deliberative process. In another case, in assessing the risk of finding a collegial colleague, the committee overtly discussed the candidate’s likability but, perhaps because members of the committee were aware that research showing issues of collegiality and fit often masks stereotypes (e.g., White-Lewis, 2020), collegiality and likability were important and acknowledged as such implicitly, but often not discussed directly, and not against the formal criteria used in the rubric. Yet, the collegiality risk was also very much institutionalized and expected, and clearly something committees had expected to find in past searches, as they did in this one.

Said differently, the identification and assessment of risks had become part of the search committees “repertoires of practice” and, as such, they had become viewed as normal terms of engagement (e.g., going faster reduces the loss of good candidates) and grounds for judgement (e.g., a candidate with a pure disciplinary focus is a safer bet than an interdisciplinary candidate). As scholars and practitioners who have participated in or led inclusive hiring workshops, we can imagine a situation wherein a committee member who participated in implicit bias training might be aware of sexist or racist stereotypes and interrupt when they emerged. Yet, we could also see the same person not questioning whether moving faster could harm inclusive hiring outcomes because of how intrinsically and legitimately it seemed tied to the professional task of sorting and hiring the best—the task at hand. Yet, equity is constrained when we lack an awareness of the ways our preferences shape risk assessments (Kahneman, 2011). Although candidates expect teaching, research, and service criteria for the positions that they applied for, and in some cases would have been socialized toward the hidden curriculum of expectations for collegiality (Liera, 2020), the risks identified by the committee and the process of assessing those risks were largely invisible and unseen by candidates. Given that the tacit preferences (e.g., candidates who had family in the area) were not transparent, candidates did not have equal access to showing how they met these preferences (Posselt, 2016). Posselt and colleagues (2020) observe that biases “maintain their power largely through their invisibility” (p. 18) and we would posit that the same is true of risks. Without knowing what search committee members considered to be compelling risks, how could candidates equitably show their strengths in those areas?

Our second point is that we found risk identification and assessment very much shaped by committee member reference points—which were affected by emotion, beliefs, disciplinary logics, and structural and implicit biases. Consistent with the cognitive bias of “anchoring,” wherein a decision-maker identifies a set point and makes decisions with it in sight (Kahneman, 2011), committee members had strong preferences that they carried into their identification and assessment of risks. Some of these reference points were part of disciplinary and prestige logics (Austin, 1990; Posselt, 2016), some were shaped by institutional type (Bell & Chong, 2010; Gonzales & Terosky, 2016), and some were hiring practices common in the field of higher education (Stewart & Valian, 2018). Each of the reference points and preferences implicitly shaped the nature of the risks and the weight given to them. For example, committee members at teaching-focused universities had greater concern with the risk of teaching competence, whereas those at research universities were concerned with the prestige and reputation of scholarly work.

Committee members also enacted implicit biases in framing and assessing risk, consistent with prior research in hiring. Like how a faculty member might see two CVs and implicitly favor the CV with a name that sounds white and/or like a man (Eaton et al., 2019), we found relationships between assessments of some risks and a candidate’s perceived social identities. For example, we found that questioning of quantitative expertise was raised more often for women candidates, and we found questions regarding English speaking were raised among international applicants of color. Research shows that racially minoritized faculty and women faculty in fields where they are underrepresented work doubly hard to prove their worth (Gutierrez y Muhs et al., 2012). In related work, scholars have found women engineers more likely to be questioned about their research topics (Blair-Loy et al., 2017). Assessments of collegiality often go badly for candidates who do not confirm to social role congruence and gender stereotypes (Cheryan & Marcus, 2020). We found questions of whether a Latina woman might be “judgey” that combined both a likability and collegiality risk, and that was not applied to other candidates who asked similar questions to the search committee. Research has shown bias in perceptions of faculty competence (Gutierrez y Muhs et al., 2012) and collegiality (Danowitz Sagaria, 2002), and against areas of study that are more likely to be populated by minoritized groups (Settles et al., 2018). We also heard a previously documented deficit narrative that faculty of color, simply by virtue of being underrepresented, would have multiple offers (Smith et al., 1996). In multiple cases, implicit biases acted as reference points that made some marginalized candidates “riskier.”

The final point is that the conditions present within faculty hiring, also present in other evaluative contexts, shaped risk assessment and made it more likely that risks would be identified with racially minoritized and women candidates. The first condition was that risk identification and assessment was almost completely unscripted, or outside of administrative oversight (Ridgeway & Correll, 2004). Each search identified formal job criteria which either an equity administrator or hiring official reviewed as part of the hiring process. However, the identification of risks and process for assessing them was largely improvised. We observed committee members taking in information as varied as a candidate smile or lack thereof, a single line on a CV, and/or information about family, and making meaning of it for a high-stakes decision without any guidance or way to do so equitably. Consistent with past research (Ridgeway & Correll, 2004), absent a script, and improvisations allowed biases to seep into risk assessment.

The second condition was that evaluators made risk assessments inconsistently in evaluating candidates. While some risks were assessed for all candidates in the negative and positive (e.g., this person does more interdisciplinary vs. pure disciplinary work), other risks were assessed differently depending on the candidate. For example, we saw committees aware of the positive benefits minoritized candidates might bring to their department, but rarely did we see committees voice a concern that a candidate would not contribute to diversity missions. Such inconsistencies are similar to past research showing the “shifting standards” that candidates from minoritized groups often encounter in evaluation (Biernat et al., 2010), and can undermine the equity and effectiveness of evaluation (Posselt et al., 2020).

The third condition relates to the ambiguity of the criteria and process the search committee members used to assess risks. Committee members struggled to assess risk with information that seemed uncertain and highly subjective (Epley & Krueger, 2005). For instance, committees exhibited insecurities about the risk of hiring a candidate who was “too good” for the department/institution. This insecurity occurred least at the research-intensive institutions and was more present at the Doctoral and Baccalaureate institutions. This subjective assessment often involved information that committees unintentionally received about one candidate. Research shows that when evaluation criteria and processes are overly vague, unclear, or ambiguous, they disproportionately hurt women and racially minoritized faculty by making it more likely that evaluators will engage social biases to make decisions (Beddoes & Pawley 2014; O’Meara et al., 2021; Stewart & Valian 2018). In this case, perceptions of candidates who might be “flight risks” activated bias because committees made assumptions about the preferences of candidates for higher ranking or more prestigious institutions.

The tendency for preferences, tastes, emotions, and reference points, to shape perceptions of risk and risk assessment is inevitable and part and parcel of the discretionary, subjective, and constructed nature of faculty evaluation (O’Meara, 2021; Posselt et al., 2020). In addition, obtaining approval to hire a new tenure track faculty is increasingly rare in many higher education institutions. Diminishing resources for departments to acquire tenure track hires may heighten risk aversion such that faculty become more conservative in selection criteria to “get it right.” While judgement and discretion can be used to reproduce inequities, they can also be employed in an agentic way, to disrupt power asymmetries, structural disadvantages, and individual biases (Gonzales & Terosky, 2016; O’Meara, 2021; Posselt et al., 2020). We turn to how that might occur through research and practice in the next section.