The structural diversity of faculty in higher education remains limited, particularly considering race and gender. Examinations of faculty hiring processes often engage deficit-based explanations rooted in the “pipeline” of potential candidates who have historically minoritized identities. Despite the presence of anecdotal beliefs about the job market for doctoral students and early career researchers, empirical evidence about their experiences, qualifications and successes remains absent from existing literature; this study aims to explore this gap.
The quantitative data for this exploratory analysis come from the 2019–2020 administration of the Job Search Collaborative Applicant Survey. The sample includes over 300 doctoral students and early-career researchers seeking faculty positions in humanities or social sciences in the USA. The authors provide descriptive data and results from regression analysis.
The results of the study provide data that informs realities of candidates’ experiences on the faculty job market in humanities and social sciences during 2020. No significant differences emerged based on race or gender across multiple productivity metrics (e.g. publications, citations). Despite this, other social identities (age, disability and first-generation status) did have significant relationships with outcomes and experiences in the faculty job market.
The continual focus on increasing candidate productivity through an emphasis on normative academic metrics (e.g. publications, citations) is not the only important consideration within the faculty job search process and may not directly yield success for doctoral students and early-career researchers in humanities and social sciences.
Introduction
Just over two decades ago, the American Association for Higher Education sponsored a large-scale study examining faculty careers and employment, with a focus on understanding experiences of graduate students and junior faculty (Rice et al., 2000). The study examined how senior faculty members’ careers looked markedly different from the career pathways for the next generation of scholars. Despite continued calls for diversifying the faculty in US colleges and universities, the argument of a limited “pipeline” of diverse candidates remains pervasive but also challenged (Griffin, 2020). For example, in a study of medical school faculty lines, Gibbs et al. (2016) developed a simulation that suggested that an exponential growth of underrepresented minorities (URM) PhD graduates and additional five decades of hiring, a continuation of existing practices without innovative interventions would still fail to substantially increase faculty diversity. Even when institutions espouse a desire for diversifying faculty, the notion of “fit” as a selection criterion for new faculty inhibited hiring candidates who would further diversify the faculty (White-Lewis, 2020). Increases in faculty diversity emerge most frequently in lecturer, adjunct or other non-tenure track roles (Griffin, 2020).
Pulling back the curtain to examine the realities of the US faculty job market in humanities and social sciences (HSS) can also provide knowledge that can inform those training and mentoring future faculty. Mentorship remains a commonly espoused recommendation for supporting minoritized doctoral students, but students often acknowledge a lack of available mentors (Fernandes et al., 2020). Thus, learning about the present realities of the faculty job market has potential for providing additional evidence that can enhance targeted support for and approaches of job seekers. With the continued pervasiveness of inequality in the compositional diversity of faculty, understanding more about the faculty job search also provides one avenue for examining structural antecedents to faculty diversity.
Using empirical data about the actual experiences of faculty job search candidates has the potential for challenging assumptions that may exist about faculty hiring in the USA. Beyond STEM, within education research, examining HSS together presents a common practice for studies (Mason and Merga, 2018; Wolfinger et al., 2008). The purpose of this study is to bring a critical perspective to learning about the HSS faculty job search candidates and their experiences during the 2020 job search cycle. Through this study, we seek to uncover equity-based realities of the faculty job market and hiring process. In this exploratory analysis, we ask:
What are the experiences, characteristics and qualifications of social sciences and humanities faculty job seekers in 2020?
Are there relationships between candidates’ qualifications and job market success?
Do differences exist between normative productivity metrics based on candidates’ social identities?
Are there relationships between candidate labor in the faculty job market (number of applications submitted, interviews conducted) and measures of job market success when considering candidates’ social identities?
Guiding literature and conceptual considerations
The historic and present realities of higher education environments hold important information regarding the environmental experiences and events that surround candidates during their socialization, preparation and faculty search processes. These experiences are deeply entangled with longstanding inequalities that impact access, opportunity and evaluation in academia. We draw from both empirical literature and the conceptual framework of person–environment (P-E) fit to inform our understanding of faculty hiring and guide this study’s analytic choices.
Historic and continual manifestations of identity-based power in higher education
Manifestations of inequality span all aspects of faculty experiences. Inequality appears in hiring, retention, promotion, funding, service responsibilities, teaching evaluations, job satisfaction, experiences with microaggressions, counteroffers, work–life integration and productivity when examining comparisons by race (Bavishi et al., 2010; Chen et al., 2022; Harris et al., 2021; Hur et al., 2017; White-Lewis et al., 2025), gender (Bavishi et al., 2010; Berheide et al., 2022; Box-Steffensmeier et al., 2015; Morgan et al., 2022; Morrison et al., 2011; O’Meara et al., 2017; White-Lewis et al., 2024; Wolf-Wendel and Ward, 2006), socioeconomic privilege (Morgan et al., 2022) and other historically minoritized identities (Social Sciences Feminist Network Research Interest Group, 2017). Extensive scholarship highlights these differential experiences for historically minoritized postdocs and faculty within higher education. In response to these pervasive inequalities, the imperative for diversifying faculty continually emerges as a policy solution advocated by students, faculty, administrators, professional associations and even many policymakers. Despite wide agreement on the importance of diversifying the faculty, substantial progress is unambiguously limited.
Metaphors of the “leaky pipeline” remain prevalent as a justification for the inability to diversify the professoriate; the argument suggests that if more diverse students had doctoral degrees, faculty diversity could increase (Griffin, 2019). This metaphor engages a false deficit perspective regarding the attrition of students with historically minoritized identities who do not advance through academia to faculty positions. Instead of acknowledging institutional responsibility, the passive conception of students “leaking” does not align with the reality that the percentage of faculty of color is not proportional with the growth of doctoral graduates of color and lacks gender parity (O’Meara et al., 2020). Conceptualizing the pipeline as “leaking” scapegoats institution practices and gatekeepers who enable the perpetuation of the leakiness and systematically facilitate it by excluding minoritized faculty from faculty positions. With this framing, strategies for diversifying the faculty “have largely focused on skill development and preparation for faculty careers” (Griffin, 2019). While these represent important foci, Griffin acknowledges the imperative of naming and examining the ways people, policies and systems maintain oppressive campus environments that limit or inhibit recruitment and retention of diverse faculty.
These manifestations of power also appear in metrics of scholarly productivity, which, again, represent influences of institutional structures, including the lack of opportunity structures for graduate students from historically minoritized backgrounds (O’Meara et al., 2020). The appearance of inequality emerges when examining many metrics, including author position (Bendels et al., 2018; Dion et al., 2018; Fox and Paine, 2019), representation and citation (Chakravartty et al., 2018; King et al., 2017; C. A. Smith et al., 2021; Thelwall, 2018, 2020) and grants (Ginther et al., 2011; Hoppe et al., 2019) . On the job market, candidates’ credentials should matter. The reality, however, is that while metrics do matter for faculty hiring in the hard sciences, they only matter up to a specific threshold (Fernandes et al., 2020). As a specific example, in a study focused on selection criteria for mechanical engineering faculty at research universities, White-Lewis et al. (2024) found that with seemingly equivalent credentials, the perceived identity of candidates was frequently a factor in identifying the most competitive candidates. The authors rightly acknowledge that prioritizing publication record first “maintains the status-quo in the assessment of academic excellence without interrogating structural constraints to achieving excellence for marginalized groups” (White-Lewis et al., 2024). Thus, not acknowledging the elements of structural inequality that influence candidate metrics perpetuates the idea that candidates are equivalent if they have the same metrics, even though consistent research suggests differential access and inequality for historically minoritized candidates.
Contexts for faculty positions and searches
The number of doctoral recipients continues increasing (National Science Foundation, 2020), and the faculty job market is oversaturated in many disciplines (Larson et al., 2014). The COVID-19 pandemic also brought a wave of hiring freezes in response to decreased enrollment (Flaherty, 2020). With shifts in both supply and demand, competition in the faculty job market has increased, while the imperative for diversifying faculty remains a consistently espoused goal. As candidates enter and navigate the job market, nearly three-quarters feel unprepared without knowledge of expectations of the process or committee (Henderson and Syed, 2016).
Participating in the faculty job market is no doubt a laborious and exhausting process for all candidates (Fernandes et al., 2020), but oppressive experiences and marginalization may make an already difficult process even more challenging for job seekers with historically minoritized identities. For example, black academics giving academic presentations (e.g. job talks) often have the added burden of developing and implementing strategies for navigating racial bias, microaggressions and stereotypes that emerge during their presentations (McGee and Kazembe, 2016). These added burdens can contribute toward racial battle fatigue, which is associated with decreased sense of belonging and overall health (Smith, 2014) that make an already difficult process increasingly challenging for racially minoritized candidates.
Looking specifically at the processes used by search committees highlights additional potential inequalities. Liera and Hernandez (2021) highlighted that committee members evaluate minoritized candidates more harshly, have different participation patterns during their interviews (e.g. not attending certain sessions for minoritized candidates) and use different criteria by constructing ideas of “merit” that are “normed to white, male, and Western European epistemological ideals” (Liera and Ching, 2019). In a study about the influence of using rubrics in searches as a tool for limiting inequitable evaluation, Culpepper et al. (2023) did identify some practices that improved diversity in hiring, but they also found that some committees built and calibrated rubrics around criteria that preserve and reinforce structural inequalities. Inequalities also emerged for international candidates, for example, committees cited concerns with “language barriers” for candidates who had non-Western accents. Their study also unveiled difficulty for committees determining how to define and measure DEI-criteria such as discussing how a committee “debated whether certain candidates, including women and international candidates, ‘counted’ as underrepresented” (Culpepper et al., 2023, p. 842). Focusing on the candidates’ skills, knowledge and potential contributions to DEI emerged only when explicit in rubric criteria; otherwise, committees focused exclusively on identities. Even when early stages of the searches included rubrics, decisions and committee conversations focused less on rubric criteria as the search advanced. These findings align with Liera and Hernandez (2021) who detailed manifestations of color-evasive racism within the later stages of search processes. Thus, even with acknowledgment of systemic issues and attempts toward eliminating them, such as hiring rubrics, the processes can still devolve as they move toward the final decisions.
Conceptual framing and variable selection
To connect this literature to our analytic approach, we draw on the framework of P-E fit, which focuses on the compatibility between individuals and institutional environments (Kristof-Brown et al., 2005). Within faculty hiring, this concept extends beyond qualifications to include alignment with departmental norms, disciplinary expectations and institutional values. However, as prior work demonstrates, perceptions of fit are not neutral. They are shaped by dominant epistemologies, norms of professionalism and culturally specific ideals that often privilege white, male and Western identities (Liera and Ching, 2019; White-Lewis et al., 2024). As such, P-E fit becomes a critical lens for understanding how inequality persists even in seemingly meritocratic evaluations.
This study applies the P-E fit framework to analyze how candidate characteristics and job search behaviors interact with institutional expectations to influence hiring outcomes. Candidates’ preparation, as measured through teaching experience, publications and grant activity, serves as a proxy for both their readiness and their access to opportunity structures. However, we interpret these metrics not as neutral indicators of merit, but as reflections of access to resources and support – often stratified by identity and institutional affiliation. Similarly, candidates’ job search intensity – captured through number of applications and interviews – offers insight into their strategic engagement with the market but also signals the degree to which they are navigating structural barriers, such as limited networks or racial battle fatigue (Kanfer et al., 2001; Smith, 2014).
Importantly, we do not treat all candidates with similar metrics as equivalent. Rather, our analytic approach acknowledges that identical credentials may be evaluated differently depending on the candidate’s identity, institutional context and how their experiences are interpreted by hiring committees. The reviewed literature suggests that these interactions are central to understanding outcomes. Our conceptual framework shown in Figure 1 reflects these dynamics, illustrating the interplay between candidate attributes, institutional gatekeeping and structural inequality. It provides a foundation for interpreting the statistical models used in this study and situates our findings within a broader conversation about equity and access in the academic job market.
The framework illustrates how productivity, preparation, and identity as individual attributes, combined with job search behaviors such as applications, interviews, and institution type, contribute to person environment fit, which in turn affects hiring decisions. Structural contexts, including institutional policies, market conditions, and resource availability, also shape personenvironment fit. Bidirectional links between individual attributes and job search behaviors indicate mutual influence.Conceptual model of the faculty job market dynamics. This model illustrates the hypothesized relationships among candidate characteristics, job search behaviors, structural and contextual influences and faculty hiring outcomes. The model positions hiring decisions as a function of who the candidate is at the time of the search, including their preparation; how they conduct their search, and the interdependence this has with who the candidate is; and job availability and the practices of the institutions they apply to. These functions are all then filtered through an person-environment fit determined by both the candidate and the search committees to end at the outcome of a job offer
Source: Authors’ own creation/work
The framework illustrates how productivity, preparation, and identity as individual attributes, combined with job search behaviors such as applications, interviews, and institution type, contribute to person environment fit, which in turn affects hiring decisions. Structural contexts, including institutional policies, market conditions, and resource availability, also shape personenvironment fit. Bidirectional links between individual attributes and job search behaviors indicate mutual influence.Conceptual model of the faculty job market dynamics. This model illustrates the hypothesized relationships among candidate characteristics, job search behaviors, structural and contextual influences and faculty hiring outcomes. The model positions hiring decisions as a function of who the candidate is at the time of the search, including their preparation; how they conduct their search, and the interdependence this has with who the candidate is; and job availability and the practices of the institutions they apply to. These functions are all then filtered through an person-environment fit determined by both the candidate and the search committees to end at the outcome of a job offer
Source: Authors’ own creation/work
Methods
Data collection and sample
A survey was designed to collect self-reported demographics and academic metrics for assistant professor applicants during the 2019–2020 academic job search cycle. The survey was open from May 6, 2020, to September 3, 2020, and respondents were not required to answer all questions. Variables of interest included faculty application outcomes such as interviews, offers, characteristics of institutions where candidates progressed and applicant demographics, including gender, race, discipline, current position (e.g. student, postdoc) and first-generation status. We examined missing data patterns using the MICE package [md.pattern()], finding minor missingness (age, 1%; first-generation undergraduate status, 1%; peer-reviewed papers, 2%; first-author papers, 4%) and substantial missingness for scholar citations (42%). Missing values were imputed using a single iteration of default mean/mode imputation [mice(df, maxit = 0)], resulting in a single completed data set.
The survey was distributed on various social media platforms, including the Future PI Slack group, Twitter and Facebook, and by several postdoctoral association mailing lists in North America, Europe and Asia. The survey was approved by the University of North Dakota’s IRB office under project number IRB-202003-240.
Data categorization and variables
We classified survey respondents who indicated that they were either non-binary, trans gender or that their gender was not listed as gender non-conforming (TGNC). Due to the low number of TGNC respondents in the data set, we included them with women to retain them within the study. We also grouped respondents – into two categories – according to their self-identification with several race/ethnicity categories. We considered respondents who self-identified with one or more of the following identities as persons excluded due to ethnicity or race (racially minoritized): Black/African/African American, Oceanic, Not Listed, North American Indigenous, Caribbean Islander, and North American Hispanic/Latinx. Non-racially minoritized respondents included those who only identified as European/Caucasian-American, North African or Middle Eastern/Caucasian-American and/or Asian/Asian American.
We removed participants who completed less than 33% of the survey questions and those who selected a disciplinary affiliation besides “Social, Behavior, and Economic Sciences” or “Humanities.” The number of survey respondents in the final data set was 276. Table 1 provides descriptive data about the full array of participants. Additionally, among participants, they had an average of 4.82 (σ = 2.80) teaching experiences (e.g. instructor of record, teaching assistant).
Participant demographic information
| Descriptor | N | Percent* |
|---|---|---|
| Age | ||
| < 30 years old | 48 | 17.4 |
| 31–35 years old | 118 | 42.8 |
| 36–40 years old | 69 | 25 |
| 41+ years old | 40 | 14.5 |
| Gender | ||
| Man | 68 | 24.6 |
| Trans/GNC | 22 | 8 |
| Woman | 175 | 63.4 |
| Racially minoritized | ||
| No | 224 | 81.2 |
| Yes | 52 | 18.8 |
| Disability | ||
| No | 217 | 78.6 |
| Yes | 50 | 18.1 |
| First-gen PhD | ||
| No | 39 | 14.1 |
| Yes | 233 | 84.4 |
| First-gen undergrad | ||
| No | 197 | 71.4 |
| Yes | 77 | 27.9 |
| Dependents | ||
| No dependents | 217 | 78.6 |
| Yes, multiple children/adult(s) | 28 | 10.1 |
| Yes, one child | 31 | 11.2 |
| Current position | ||
| Non-tenure track faculty | 66 | 23.9 |
| PhD candidate (ABD) | 63 | 22.8 |
| Postdoc | 108 | 39.1 |
| Discipline | ||
| Humanities | 109 | 39.5 |
| Social, behavior and economic sciences | 167 | 60.5 |
| Residence | ||
| Canada | 34 | 12.3 |
| Other | 13 | 4.7 |
| USA | 228 | 82.6 |
| Fellowship | ||
| Yes | 116 | 58.9 |
| No | 81 | 41.1 |
| Social media use | ||
| NA | 17 | 8.6 |
| No | 60 | 30.5 |
| Yes | 120 | 60.9 |
*Respondents were able to skip questions they were uncomfortable answering, within question responses may not equal 100%
We used R statistical software (version 4.3.2) and related packages for data manipulation and visualization and relevant packages. For regression models, we included participants who indicated their application outcome. All code used for data analysis and visualization are available in the GitHub repository: Link to the website of GitHub..
Analytic approach
After reviewing the descriptive statistics, three separate logistic regression models were used to examine the relationships between candidates’ social identities, job market labor, academic qualifications and job market success, operationalized as receiving a job offer for any tenure-track or tenure track equivalent. All models used the “glm” function in R with a binomial family to account for the binary outcome variable (job offer, yes/no). Predictor variables were standardized using the scale() function to facilitate comparison across different variables. Specifically, Model 1 examined the impact of age, disability status, race, gender, first-generation undergraduate status and legal employment status (i.e. employment visa needed) on job market success. Model 2 extended Model 1 by including on-site interviews as a measure of candidate labor to consider potential tokenization associated with seeking diverse pools at all stages of hiring. Model 3 extended Model 1 by including scholarly citations as a measure of academic qualifications.
Limitations
There are limitations associated with the present research study. These data support examination of several aspects of the faculty search process for candidates in social sciences and humanities. The sample size represents a convenience sample of individuals who were aware of and completed the survey, which may have captured participant data that does not support generalizability to those who did not complete the survey. A larger sample size would support more opportunities for disaggregating academic disciplines to provide more nuanced understanding. Additionally, as a cross-sectional survey, we do not have a longitudinal understanding of how candidates’ job search progressed beyond the timeframe covered in their survey response. Some variables within the study had substantial missingness. In particular, scholar citations, used in Model 3, included more missingness, and thus, the findings are more tentative. Additionally, the data do not specify the type of positions (e.g. visiting faculty, contingent, tenure-track) for applications or outcomes, which may have contributed toward more nuanced understanding. The item on participants’ fellowship experiences did not request distinctions between graduate fellowships and postdoctoral fellowships; future studies should allow this clarification from participants. Finally, our exploratory approach to Model 2 is limited due to the high correlation of participating in an onsite visit and receiving an offer; our data for this exploration also does not support analysis with continuous variables to gather nuance on the yield of job offers related to the number of onsite visits.
Results
The descriptive analyses associated with the first research question begins providing evidence that frames the realities of job seekers in HSS. More than 75% of survey participants had some form of fellowship. While nearly all participants had some teaching experience, an overwhelming number had experience beyond serving as a TA. Participants in the humanities participated in significantly (p < 0.01) more application cycles. Candidates in the humanities had lower productivity metrics across all measured areas. Job seekers in HSS engaged social media and web presence differently as well. Half of the candidates in social sciences had a blog or website, which was significantly (p < 0.05) more than humanities where just a quarter of participants had a blog or website. Participants in the social sciences also reported significantly (p < 0.05) more social media presence. Figure 2 provides further details about the 2020 job seekers in the study.
The grouped charts present respondent data across multiple dimensions of academic job applications. Panel (a) shows the number of application cycles among respondents. Panel (b) depicts the number of postdoctoral positions held. Panel (c) summarises research output metrics such as first-author and peer-reviewed papers, citation counts, and scholar indices. Panel (d) illustrates different types of teaching experience, including assistant, adjunct, and certificate-based roles. Data are presented as percentages across humanities and social, behavioral, and economic sciences, highlighting variation in publication records, research metrics, and teaching experiences among academic applicants.Comparison of applicant experience and metrics by discipline. Those respondents who self-selected as belonging to either the humanities (blue) or the social, behavioral and economic sciences (gray) were pulled from all respondents to the 2019–2020 faculty job market survey. Responses to survey questions were analyzed according to the percent of respondents in each field. The number of (a) application cycles and (b) postdoctoral positions participated in as of the 2019–2020 cycle. (c) Boxplots and medians of within-field applicant metrics: (top left) first-author papers, (top right) Google Scholar citations, (bottom left) Google Scholar h-index and (bottom right) peer-reviewed papers. (d) Analysis of teaching experience by field: (top) relative to teaching assistantships (TAs) and (bottom) specific teaching opportunities. The Mann–Whitney U test with continuity corrections was used for statistical analysis of Panel A
Source: Authors’ own creation/work
The grouped charts present respondent data across multiple dimensions of academic job applications. Panel (a) shows the number of application cycles among respondents. Panel (b) depicts the number of postdoctoral positions held. Panel (c) summarises research output metrics such as first-author and peer-reviewed papers, citation counts, and scholar indices. Panel (d) illustrates different types of teaching experience, including assistant, adjunct, and certificate-based roles. Data are presented as percentages across humanities and social, behavioral, and economic sciences, highlighting variation in publication records, research metrics, and teaching experiences among academic applicants.Comparison of applicant experience and metrics by discipline. Those respondents who self-selected as belonging to either the humanities (blue) or the social, behavioral and economic sciences (gray) were pulled from all respondents to the 2019–2020 faculty job market survey. Responses to survey questions were analyzed according to the percent of respondents in each field. The number of (a) application cycles and (b) postdoctoral positions participated in as of the 2019–2020 cycle. (c) Boxplots and medians of within-field applicant metrics: (top left) first-author papers, (top right) Google Scholar citations, (bottom left) Google Scholar h-index and (bottom right) peer-reviewed papers. (d) Analysis of teaching experience by field: (top) relative to teaching assistantships (TAs) and (bottom) specific teaching opportunities. The Mann–Whitney U test with continuity corrections was used for statistical analysis of Panel A
Source: Authors’ own creation/work
Figure 3 illustrates candidates’ experiences in the number of applications submitted, interviews and job offers according to gender and racially minoritized identity. For both gender and racially minoritized identity, no significant differences emerged across any of the categories of applications submitted or outcomes.
The first section (a) displays box plots for applications submitted, off-site interviews, on-site interviews, and faculty offers across gender categories: man, woman or transgender or gender non-conforming, and no response. The second section (b) compares the same outcomes by P E E R identity, distinguishing between those who identify as P E E R and those who do not. Each plot shows the number of applications, interviews, and offers per applicant, highlighting distributional differences and outliers across demographic groups.Comparison of applicant participation and outcomes by gender- or racially minoritized-identity status. Boxplots of applicant outcomes, which include the number of (top-left panels) applications submitted, (top-right) off-site interviews, (bottom-left) on-site interviews and (bottom-right) faculty offers. (a) Respondents were grouped according to their self-selected gender identities, man (green) or woman/trans/gender non-conforming (orange). Gender-identity non-respondents are also shown (gold). (b) Respondents were grouped according to self-selected racial/ethnic-identities as having either a racially minoritized (yes, blue) or non-racially minoritized (no, yellow) identity. The Mann–Whitney U test with continuity corrections was used for within-panel statistical analyses; no significant results were found
Source: Authors’ own creation/work
The first section (a) displays box plots for applications submitted, off-site interviews, on-site interviews, and faculty offers across gender categories: man, woman or transgender or gender non-conforming, and no response. The second section (b) compares the same outcomes by P E E R identity, distinguishing between those who identify as P E E R and those who do not. Each plot shows the number of applications, interviews, and offers per applicant, highlighting distributional differences and outliers across demographic groups.Comparison of applicant participation and outcomes by gender- or racially minoritized-identity status. Boxplots of applicant outcomes, which include the number of (top-left panels) applications submitted, (top-right) off-site interviews, (bottom-left) on-site interviews and (bottom-right) faculty offers. (a) Respondents were grouped according to their self-selected gender identities, man (green) or woman/trans/gender non-conforming (orange). Gender-identity non-respondents are also shown (gold). (b) Respondents were grouped according to self-selected racial/ethnic-identities as having either a racially minoritized (yes, blue) or non-racially minoritized (no, yellow) identity. The Mann–Whitney U test with continuity corrections was used for within-panel statistical analyses; no significant results were found
Source: Authors’ own creation/work
Further examining the experiences of participants, Figure 4 provides descriptives, according to gender, of the distribution of candidate applications by institution type, such as research-intensive and primarily undergraduate institutions. No significant differences emerged between target institutions and gender.
The chart presents the distribution of applications submitted by respondents at predominately undergraduate institutions and research-intensive or extensive institutions. Bars represent gender categories: man, woman or transgender or gender non-conforming, and no response. The vertical axis shows the number of applications submitted in increasing intervals from 0 to 199, while the horizontal axis shows the number of responses. Data indicate that most applicants submitted between 5 and 14 applications, with men and those who did not disclose gender showing higher frequencies across categories.Comparison of target institutions by applicant gender. Applicants indicated the number of applications that they submitted to (left panel) primarily undergraduate (PU) and/or (right panel) research-intensive (RI) institutions. These responses were binned and plotted as the percent of respondents according to the institution type and self-selected applicant gender [man (green), woman/trans/GNC (orange), non-respondent (gold)]
Source: Authors’ own creation/work
The chart presents the distribution of applications submitted by respondents at predominately undergraduate institutions and research-intensive or extensive institutions. Bars represent gender categories: man, woman or transgender or gender non-conforming, and no response. The vertical axis shows the number of applications submitted in increasing intervals from 0 to 199, while the horizontal axis shows the number of responses. Data indicate that most applicants submitted between 5 and 14 applications, with men and those who did not disclose gender showing higher frequencies across categories.Comparison of target institutions by applicant gender. Applicants indicated the number of applications that they submitted to (left panel) primarily undergraduate (PU) and/or (right panel) research-intensive (RI) institutions. These responses were binned and plotted as the percent of respondents according to the institution type and self-selected applicant gender [man (green), woman/trans/GNC (orange), non-respondent (gold)]
Source: Authors’ own creation/work
When examining if candidates’ social identities influenced their success on the faculty job market (i.e. getting at least one faculty job offer), as shown in Figure 5, the logistic regression analysis suggests that certain aspects of candidates’ social identities do have a statistically significant influence on their job market success. Specifically, candidates who are younger [age: odds ratio (OR) = 0.66, p = 0.032], have a disability (disability status: OR = 1.46, p = 0.039) or are first-generation undergraduates (first-gen undergrad: OR = 1.47, p = 0.036) show a significant association with receiving a job offer.
The forest plot presents odds ratios and confidence intervals for six predictors: first-generation undergraduate, disability status, race P E E R, gender, legal status, and age. The horizontal axis represents odds ratios. Predictors with odds ratios above one, such as first-generation undergraduate and disability status, have significant positive associations with the outcome. Age shows a significant negative effect with an odds ratio of 0.66 and a p-value of 0.032, indicating decreased likelihood of the outcome with increasing age. Other predictors, including race, gender, and legal status, show no significant effects.Social identity effects on job market success. This forest plot displays the odds ratios and 95% confidence intervals for the associations between various social identities and job market success, as determined by receiving a job offer. First-generation undergraduate status (OR = 1.47, p = 0.035) and disability status (OR = 1.46, p = 0.039) significantly associated with receiving a job offer, while age shows a significant negative effect (OR = 0.66, p = 0.032). Other social identities, including race, gender and legal status, were not statistically significantly related with job market outcomes
Source: Authors’ own creation/work
The forest plot presents odds ratios and confidence intervals for six predictors: first-generation undergraduate, disability status, race P E E R, gender, legal status, and age. The horizontal axis represents odds ratios. Predictors with odds ratios above one, such as first-generation undergraduate and disability status, have significant positive associations with the outcome. Age shows a significant negative effect with an odds ratio of 0.66 and a p-value of 0.032, indicating decreased likelihood of the outcome with increasing age. Other predictors, including race, gender, and legal status, show no significant effects.Social identity effects on job market success. This forest plot displays the odds ratios and 95% confidence intervals for the associations between various social identities and job market success, as determined by receiving a job offer. First-generation undergraduate status (OR = 1.47, p = 0.035) and disability status (OR = 1.46, p = 0.039) significantly associated with receiving a job offer, while age shows a significant negative effect (OR = 0.66, p = 0.032). Other social identities, including race, gender and legal status, were not statistically significantly related with job market outcomes
Source: Authors’ own creation/work
Next, we examined if candidates’ social identities differentially influenced their labor in the job market to secure a faculty offer. As shown in Figure 6, on-site interviews were statistically significantly associated with an increased likelihood of receiving a job offer (OR = 10.61, p < 0.001), while being a first-generation undergraduate also shows a significant positive association (OR = 1.63, p = 0.043), with other social identities not showing significant associations.
This forest plot illustrates the odds ratios and confidence intervals for seven predictors: on-site interviews, first-generation undergraduate, disability status, gender, race P E E R, legal status, and age. The odds ratio for on-site interviews is 10.62 with a p-value of 0.000, indicating a strong and significant association with the outcome. First-generation undergraduate status also shows a significant but smaller effect, with an odds ratio of 1.63 and a p-value of 0.043. The remaining variables, including disability status, gender, race, legal status, and age, exhibit odds ratios below one and are not statistically significant.Effects of on-site interviews and social identities on job market success. This forest plot displays the odds ratios and 95% confidence intervals for the associations between on-site interviews and social identities with job market success, as determined by receiving a job offer. On-site interviews were significantly associated with receiving a job offer. (OR = 10.61, p < 0.001), while being a first-generation undergraduate also shows a positive association (OR = 1.63, p = 0.043), with other social identities not showing significant associations
Source: Authors’ own creation/work
This forest plot illustrates the odds ratios and confidence intervals for seven predictors: on-site interviews, first-generation undergraduate, disability status, gender, race P E E R, legal status, and age. The odds ratio for on-site interviews is 10.62 with a p-value of 0.000, indicating a strong and significant association with the outcome. First-generation undergraduate status also shows a significant but smaller effect, with an odds ratio of 1.63 and a p-value of 0.043. The remaining variables, including disability status, gender, race, legal status, and age, exhibit odds ratios below one and are not statistically significant.Effects of on-site interviews and social identities on job market success. This forest plot displays the odds ratios and 95% confidence intervals for the associations between on-site interviews and social identities with job market success, as determined by receiving a job offer. On-site interviews were significantly associated with receiving a job offer. (OR = 10.61, p < 0.001), while being a first-generation undergraduate also shows a positive association (OR = 1.63, p = 0.043), with other social identities not showing significant associations
Source: Authors’ own creation/work
After including the on-site interview as a factor, we still observe that being a first-generation undergraduate positively influences job offer outcomes (OR = 1.63, p = 0.043). On-site interviews emerged as a strong predictor of success (OR = 10.61, p < 0.001), underscoring the critical role of direct candidate engagement with employers. Other social identities like age, disability status, race, gender and legal status were not statistically significantly associated with job market success.
Finally, we examined the influence of social identities on the correlation of scholarly citations in predicting job market success. As shown in Figure 7, the analysis reveals that first-generation undergraduate status significantly enhances job market success (OR = 1.51, p = 0.027), while age continues to show a negative effect (i.e. younger individuals are more likely to receive a job offer; OR = 0.65, p = 0.027). Scholarly citations, however, were not statistically significantly related to receiving a job offer (OR = 0.77, p = 0.184), suggesting that traditional academic qualifications may play a limited role compared to social identities.
The forest plot displays odds ratios and confidence intervals for seven predictors: first-generation undergraduate, disability status, race P E E R, gender, legal status, Google Scholar citations, and age. The horizontal axis represents odds ratios. First-generation undergraduate status shows a significant positive effect with an odds ratio of 1.51 and a p-value of 0.027, while age shows a significant negative effect with an odds ratio of 0.65 and a p-value of 0.027. Disability status approaches significance with an odds ratio of 1.44 and a p-value of 0.052. Other predictors, including race, gender, legal status, and Google Scholar citations, have odds ratios near one and are not statistically significant.Relationships of social identities and scholarly citations with job market success. This forest plot illustrates the odds ratios and 95% confidence intervals for the associations between social identities and scholarly citations with job market success. The analysis reveals that first-generation undergraduate status significantly enhances job market success (OR = 1.51, p = 0.027), while age continues to show a negative effect (OR = 0.65, p = 0.027). Scholarly citations, however, were not significantly associated with receiving a job offer (OR = 0.77, p = 0.184), suggesting that traditional academic qualifications may play a limited role compared to social identities
Source: Authors’ own creation/work
The forest plot displays odds ratios and confidence intervals for seven predictors: first-generation undergraduate, disability status, race P E E R, gender, legal status, Google Scholar citations, and age. The horizontal axis represents odds ratios. First-generation undergraduate status shows a significant positive effect with an odds ratio of 1.51 and a p-value of 0.027, while age shows a significant negative effect with an odds ratio of 0.65 and a p-value of 0.027. Disability status approaches significance with an odds ratio of 1.44 and a p-value of 0.052. Other predictors, including race, gender, legal status, and Google Scholar citations, have odds ratios near one and are not statistically significant.Relationships of social identities and scholarly citations with job market success. This forest plot illustrates the odds ratios and 95% confidence intervals for the associations between social identities and scholarly citations with job market success. The analysis reveals that first-generation undergraduate status significantly enhances job market success (OR = 1.51, p = 0.027), while age continues to show a negative effect (OR = 0.65, p = 0.027). Scholarly citations, however, were not significantly associated with receiving a job offer (OR = 0.77, p = 0.184), suggesting that traditional academic qualifications may play a limited role compared to social identities
Source: Authors’ own creation/work
When including Google Scholar citations in the model, the analyses suggest that certain social identities differentially influence how qualifications affect candidates receiving jobs. Being a first-generation undergraduate consistently emerges as a significant positive factor (OR = 1.51, p = 0.027), like previous analyses, indicating that this identity enhances job market success. Age again shows a negative effect (OR = 0.65, p = 0.027), suggesting older candidates may be at a disadvantage. Disability status approaches significance (OR = 1.44, p = 0.052), hinting at a potential positive influence. However, the effect of scholarly citations, a key qualification, was not associated with job offers (OR = 0.77, p = 0.184), nor did other social identities like race/ethnicity, gender or legal status. This indicates that while some social identities like being a first-generation undergraduate or age may be associated with job outcomes, traditional academic qualifications might not be as decisive in this context.
We also tested interaction effects to determine whether scholarly productivity moderates minoritized candidates’ likelihood of receiving a job offer, when retaining all other variables from the previous model. Results indicated a significant positive interaction between race/ethnicity and total peer-reviewed papers (OR = 1.37, p = 0.020), suggesting minoritized candidates benefit more from publishing additional papers than majority candidates. However, a significant negative interaction was found between race/ethnicity and first-author papers (OR = 0.51, p = 0.049), indicating that first-author publications do not confer the same advantage for minoritized candidates as for majority candidates. Additionally, scholar citations showed no significant interaction (OR = 0.99, p = 0.855), suggesting that higher citation counts alone do not disproportionately benefit minoritized applicants. Overall, these findings challenge the notion that minoritized candidates simply need equal or greater scholarly productivity to receive preferential consideration, highlighting the complexity of how productivity measures do (and do not) translate into job market outcomes.
Discussion
The results from the present study provide some interesting insights that continue exposing presumptions of faculty hiring as a meritocratic process. Although the study participants were navigating the job market during the precarity of the COVID-19 pandemic, which included canceled searches and revised search processes for institutions and candidates (Kozik et al., 2024), these results have value for consideration beyond the pandemic. Despite the disruptions during the 2019–2020 job cycle, studies suggest recovery of the faculty job market beginning with the 2021 cycle (Kozik et al., 2024). Previous studies consistently identified significant differences in productivity metrics based on race and gender (Bendels et al., 2018; Chakravartty et al., 2018; Fox and Paine, 2019; Hoppe et al., 2019; Thelwall, 2020). Within the present sample of HSS job seekers, we did not find significant differences by gender or race/ethnicity for publications, citations, author order, grants or h-index.
Multiple potential explanations exist that could contribute to this shift. Optimistically, it is possible that some grass-roots initiatives, such as #citeasista (citeasista.com), which black women doctoral students started to increase citation networks, are shifting productivity metrics and h-index in systematic ways. Despite the lack of significant differences in productivity metrics when comparing by gender or racially minoritized identity, we caution assumptions that these data imply equity. Rather, we contend that historically minoritized candidates developed tools and strategies that challenge oppressive environmental events, practices and norms. Similar findings were reported in a qualitative study of PhD students of color on the academic job market during the COVID-19 pandemic (Liera and Rodgers, 2025). Specifically, they acknowledged that despite the challenges, their participants brought strength to the ways they engaged power and agency in the search. White-Lewis et al. (2024) further caution against categorizing comparable candidate metrics in isolation within search processes because the experiences of candidates in achieving those metrics were not isolated from the oppressive contexts of the academy. We support their contention that while metrics should matter, they must be considered holistically instead of viewed “in isolation…and applied as blocks akin to building a pyramid: productivity as the base, other qualifications throughout, and identity as the tipping point when all things are considered equal” (White-Lewis et al., 2024). Taking a combined approach to consideration is particularly important during times of societal challenges, such as those experienced by the participants in the present study who were navigating the COVID-19 pandemic.
Looking at where applicants submitted applications also has value. Assumptions exist that teaching institutions have environments that are more conducive to family/personal life (Berheide et al., 2022). These assumptions also extend to a suggestion that perhaps women are overrepresented in teaching institutions because they self-select and choose institutions that are not research intensive. Our analysis in Figure 3, which showed no significant relationship between gender and institutional type for applications, again suggests that the pipeline of applicants may not be the limitation to diversifying hiring by gender. Seeing a lack of significant differences affirms that applicants may not be self-selecting out of opportunities. Perhaps, these data indicate a growing awareness of research which affirms that women can succeed with positive work/life integration even at research-intensive institutions (Berheide et al., 2022; Wolf-Wendel and Ward, 2006).
The continued presence of inequalities across candidates’ credentials could, and often does, facilitate a potential route for hiring committees to justify their biases without critical consideration of the structural realities that influence seemingly “objective” metrics for evaluation (O’Meara et al., 2020; Sensoy and Diangelo, 2017; White-Lewis et al., 2024). Within the present study, without significant differences across many productivity metrics, subjective assessments of candidates’ scholarship may represent one way that power differentiates candidates’ research even when no significant differences emerged regarding quantity. White-Lewis (2020) emphasized the ways committee members assessed candidates’ scholarly contributions, often categorizing research focused on historically minoritized communities as too narrow or lacking substantial potential for impact. Thus, although productivity metrics did not differ significantly, the systemic norms within departmental hiring practices may still have resulted in inequitable evaluation of candidates’ scholarly contributions.
Even without significant variation across productivity metrics, these results demonstrate inequality in hiring practices based on candidates’ social identities. While race and gender are commonly discussed as contributing to inequality, this paper uniquely identified the influence of disability and first-generation status. It is possible these data about first-generation status are early evidence of the use of first-generation undergraduate status as a proxy for diversity in faculty hiring, which has been previously applied in selective admissions (see Hill et al., 2023). More than 50% of first-generation students are individuals with historically minoritized racial and ethnic identities (Schuyler et al., 2021). In a time in the USA where words like “race” and “diversity” are facing federal challenges for elimination (National Center for Science and Engineering Statistics (NCSES), 2025; Palmer, 2025), perhaps these findings suggest faculty hiring in the HSS may already be on a helpful path of hiring faculty who were the first in their family to attend college.
As doctoral candidates and early-career job seekers engage in the faculty job market, the results from this study can help illuminate realities of a process that has too long remained enigmatic. When considering where to put labor, understanding factors that do – and do not – have a relationship with successful job outcomes can help job seekers and doctoral students prioritize their energy in preparation for and on the job market. These results also have utility for those conducting faculty searches to more intentionally examine potential biases that may be influencing their hiring or conceptions of “fit.”
Implications for practice
While these results suggest the potential for some shifts within the oppressive status quo in some areas (e.g. citation counts, productivity metrics), much work remains toward dismantling and reimagining longstanding processes and norms of faculty hiring. Using some of the pillars of Linder’s (2019) power-conscious framework as a guide, we provide recommendations for practice that we believe have potential for shifting the environmental events that shape faculty job search processes. Bettencourt et al. (2021) demonstrated the applicability of the framework with their development of a power-conscious model of doctoral mentorship and socialization. Many of their reflective questions could also provide valuable starting places for dialogue and action for search committees. We encourage faculty to explore questions such as these:
What are my implicit and explicit expectations of [faculty candidates]? How and where did these come from?
In what ways can I create structures of accountability with my colleagues to collectively call attention to existing structures?
How are the voices of [diverse colleagues] integrated into [our search process]? (Bettencourt et al., 2021).
Much like this paper begins exposing realities of the search process, committees must interrogate their environments to expose oppressive practices, policies and norms.
Engaging conversations and dialogues like these on a regular basis as part of normative culture in departments would also potentially provide an acceptable framework that moves beyond training programs, which are increasingly facing bans in many US states and federally. Instead of just providing revised policies or rubrics, departments and search committees must engage collective dialogue and arrive at shared understandings. If left to engage on their own, members may choose not to read new standards, instead operating based on prior precedent, and/or may arrive at varying understanding for how to think about rubrics (Culpepper et al., 2023; Liera and Hernandez, 2021). Providing space for collective conversation, in which all colleagues’ voices are valued, is critical for advancing these practices in ways that can make a positive contribution towards increased equity in faculty hiring.
Mounting challenges to these conversations should not preclude committees from engaging them. Instead, where prohibited by law from expressly discussing social identities in connection with hiring, committees can – and should – still examine the history and context of the faculty line, program, department and college. Current US laws do not prevent interrogating questions such as “What are my implicit and explicit expectations of faculty candidates? How and where did these come from?” Instead, faculty must engage their critical consciousness to inform holistic responses to these questions that align with institutional missions and professional expectations from disciplinary associations – many of which still prioritize increasing equity in hiring.
Providing critical consciousness and self-awareness conversations for search committees, faculty and administrators would require intentional self-examination of individual power, privilege and oppression associated with social identities but also the positional power ascribed to hiring committees. Looking at history and context must include not only the extensive history of oppressive policies, practices and climates but also the specific history and context of the faculty line. How might ideas and preconceptions about what the line or positions “should be” influence hiring decisions? Instead, hiring committees must challenge their normative approaches and examine how the influences of power through historical precedent may be influencing their conception of who does (or does not) “fit” for a position.
Implications for research
Results from this study highlight many areas of importance for continued research toward understanding the realities of the faculty job market and initiatives focused on diversifying faculty. Pulling back the curtain on HSS candidates’ experiences on the faculty job market has uncovered incongruencies between espoused rhetoric for candidate success and the realities of candidates with diverse identities.
The results from this study have generated data to understand assumptions about the faculty job market for HSS candidates, but scholars must continue expanding empirical knowledge about the hiring processes as a tool for dismantling the oppressive structures of faculty hiring. Data associated with the present study precluded examination of the influences of participants’ socioeconomic backgrounds. Given the results from Morgan et al. (2022) highlighting substantial differences by socioeconomic background, future studies should include that variable and focus on understanding the unique influences of socioeconomic background and race on the faculty job market. Further, disaggregating candidates’ research background beyond just HSS could support identification of additional disciplinary-specific patterns. While a lack of structural diversity remains a challenge across higher education, disciplinary differences do exist by field regarding representation and productivity metrics (Ceci et al., 2014; Schmitt, 2015). The innovative modeling conducted by Gibbs et al. (2016) focused on hiring in medical schools presents another important opportunity to explore simulations of the pathways into the professoriate for HSS candidates. They acknowledged a substantial departure of PhDs who aspired toward non-faculty careers. Their model then explored the potential increase in URM hiring by broadening the pathways to reengage potential faculty candidates who initially aspired to other careers. While applicable to medical school hiring, the prevalence of postdoctoral opportunities in HSS remains limited and the percentage of HSS PhD graduates employed in academe remains much higher than STEM counterparts (National Center for Science and Engineering Statistics (NCSES), 2025). Although the career pathways for HSS graduates may look different, Gibbs and colleagues’ models highlight the importance of using research to continue reimagining pathways and interventions to enhance faculty diversity.
Continuing the inclusion of empirical evidence within discussions and initiatives focused on supporting candidates in the faculty job must continue. Engaging with and looking at work using Scopus and bibliometrics data offer novel databases for continuing this critical research. Extending beyond just the hiring process and examining the full career trajectory for early-career faculty has potential for illuminating roots of additional systematic inequalities within academia. In an examination of more than 30,000 tenured and tenure-track faculty, Webber and Rogers (2018) found gendered differences in faculty job satisfaction across multiple factors, including institutional type and institutional control; significant variations also emerged when examining race. Again, such studies highlight systemic inequalities of faculty experiences within academia and while institutional changes must happen, centering empirical knowledge as a focus of the faculty job search process should happen concurrently with movement towards larger structural changes. Moving beyond assumptions and anecdotal beliefs about the job market and faculty life provides an important tool in support of doctoral students’ socialization, mentoring and agency within the process. Bridging scholarship that represents the trajectory of faculty careers has potential for supporting candidates in navigating the job search process and hopefully finding institutions where they feel holistically satisfied and affirmed.
Conclusion
Despite the growing body of research focused on faculty job search processes, many assumptions and anecdotes exist regarding experiences and realities of the process. Using empirical data to bring light to the realities of candidates’ experiences in the HSS faculty job search provides important evidence that can begin challenging incorrect – and oftentimes harmful and oppressive – assumptions and beliefs. This exploratory paper made strides toward examining the distinct realities for candidates outside of STEM disciplines by using critical approaches to understand potential relationships between systemic inequality in higher education and the experiences of candidates navigating the faculty job market during the COVID-19 pandemic. Several of the findings present opportunities for hope and optimism regarding potential shifts in candidate agency and hiring practices. Still, additional research is needed to continue challenging notions of a “leaky pipeline” and instead bring light to more expansive pathways that can support success for diversifying HSS faculty. So too, continued intentionality must exist across all phases of hiring processes in ways that center equity and continue to expose the opaque realities that emerge behind closed doors where final ranking and hiring decisions are made.
