Introduction

Voluntary employee turnover is a natural feature of large organizations. Although much energy is spent trying to prevent it, some amount of turnover mitigates employee stagnation and dissatisfaction that negatively impacts organizational performance, and it invites new entrants who can bring new and beneficial perspectives (Hom & Griffeth, 1995; Hom & Kinicki, 2001). In college and university workplaces, the same tenets apply to faculty, even though tenure-track faculty stay at their institutions for longer compared to employees in other industries (Kaminski & Geisler, 2012). A variety of factors influence their decisions to leave, including dissatisfaction (Daly & Dee, 2006; Johnsrud & Rosser, 2002; Rosser, 2004; Smart, 1990), compensation (O’Meara et al., 2016; Xu, 2008; Zhou & Volkwein, 2004), familial factors and geography (O’Meara et al., 2014), employer reputation and prestige (Matier, 1990; O’Meara et al., 2014), subtle and overt forms of discrimination (Matier, 1990; Zambrana, 2018) or desire for professional growth into university administration, government, or research careers outside of higher education (Barnes et al., 1998; Dorenkamp & Weib, 2018; Ryan et al., 2012; White-Lewis et al., 2022).

Though many faculty may intend to leave, most do not (O’Meara et al., 2014; White-Lewis et al., 2022; Wohrer, 2014). Results from The Faculty Survey administered by the Higher Education Research Institute (HERI) found that nearly half of 20,771 full-time undergraduate teaching faculty respondents had “considered leaving this institution for another,” in the past year (Stolzenberg et al., 2019, p. 37), but so many faculty never actually quit. Additionally, a different study of tenure-track faculty found that even among faculty who actively searched for outside offers — i.e., comparable offers of employment at another organization — half remained employed at their institutions one year later (Benson et al., 2016).

Counteroffers, wherein a faculty member receives an outside offer, but then also receives an offer from their current institution for improved salary, benefits, and/or conditions to retain the individual, are a common practice in higher education to preempt departures. When faculty present outside offers to their department(s), it signals that they have the intent to leave and the means to do so. Whether higher pay, greater research support, changes in primary job functions, or something else, the contents of counteroffers from the retaining institution typically address aspects of the job that prompted the intent to leave in the first place (O’Meara et al., 2014).

Though the norm of counteroffers is pervasive in the academy, it can be complicated in practice; the processes of “retention offers” in higher education are not well-documented. How urgently do academic leaders respond to their faculty member’s outside offers, if at all? Which outside offers are taken more seriously than others? And which faculty are pursued most heavily in counteroffers, or assumed to be unretainable? These unanswered questions show that the actual conduct of counteroffers is less linear than the extant research presents. There is no guarantee that academic leaders will feel compelled by outside offers from certain institutions or organizations, make competitive counteroffers, or do so expediently. In practice, counteroffers resemble other foggy evaluative areas in higher education such as hiring, promotion, and tenure (O’Meara, 2021). That is, faculty and academic leaders have the freedom to exercise discretion in spaces that positively or negatively impact others’ careers. But racial and gender disparities thrive in discretionary spaces, especially when they lack appropriate structure and oversight (O’Meara, 2021). The process of requesting a counteroffer, and the process of a chair or dean deciding whether and how much to offer occurs in a human system that is typically not transparent, clear, consistent, or accountable. A lack of such equity-minded guardrails (e.g., clear policies and criteria for decision-making; O’Meara, 2021) makes it likely that inequities will occur, especially for minoritized groups. Without intending to do so, deans and department chairs may inadvertently push out faculty whom they wanted to retain by misunderstanding or mischaracterizing their expressions of an intent to leave.

In this paper, we explore who is most likely to receive a counteroffer, if there are any salient differences in subsets of the faculty with respect to receiving a counteroffer, and whether there are any clear predictors of a counteroffer. Specifically, we interrogate whether there are faculty characteristics or other aspects of the context surrounding the outside offer that shape the likelihood of counteroffers. Although there are a few studies on outside offers (e.g., O’Meara et al., 2014, 2017), there is no research that examines actual counteroffers received or denied to faculty. This is especially pressing given that the process of counteroffers is largely individualized. Such confidential personnel matters have limited transparency, making their apportionment subject to less scrutiny (e.g., less oversight, uniformity, and predictability), therefore vulnerable to issues of fairness. This research seeks to shed light on the rarely examined subject of faculty counteroffers, and in doing so provide recommendations to improve leadership training on counteroffer response and negotiation processes. In what follows, we chart the departure terrain, and the subsequent climate that might precipitate a counteroffer, by reviewing its most fundamental aspects.

Literature review on outside offers and counteroffers

Faculty mobility has been studied for decades to understand the reasons why faculty leave (Matier, 1990; O’Meara et al., 2014; Rosser, 2004; Smart, 1990; White-Lewis et al., 2022; Zhou & Volkwein, 2004). Explanations run the gamut and are typically understood as either “push” or “pull” factors (Matier, 1990; O’Meara et al., 2016). Push factors are conditions of the current university that compel faculty to search for outside employment, including “internal, intangible (e.g., autonomy, satisfaction with fit) and tangible (e.g., wages, facilities) influences” (Kim et al., 2013, p. 248). In contrast, pull factors are factors external to the current university that entice faculty to relocate (Matier, 1990). These are conditions that faculty perceive they cannot presently realize, such as preferred geographical location, higher pay, and more prestige (O’Meara et al., 2016). The challenge for academic leaders is to limit the number of negative push factors and counteract pull factors with new or adjusted incentives in counteroffers. In this review, we discuss the state of research on counteroffers, including research on the outside offers that precede them, while being especially mindful of documented racial and gender inequities within these processes.

Research on outside offers

An outside offer is a comparable offer of employment at another organization (Kahn, 1998; Maertz & Griffeth, 2004). Research in this area has focused on what individual-level forces prompt an outside offer (Maertz & Griffeth, 2004), their contributions to pay inequality over time (Blackaby et al., 2005; O’Meara et al., 2017), and their differential impacts on the subsequent negotiation process (Lipp et al., 2023). Research indicates that the first offer (or outside offer) significantly shapes the playing field for subsequent negotiations and counteroffers (Lipp et al., 2023). It does so because an outside offer sends a more urgent signal than other signals (e.g., general dissatisfaction) because they indicate that the employee may have a desire to leave and has the means to do so. This is incredibly important in the higher education context compared to other sectors since many faculty suggest they want to leave but most do not (Benson et al., 2016; Stolzenberg et al., 2019), likely due in part to increasingly precarious labor market conditions for tenure track faculty positions.

In higher education, outside offers similarly represent a material desire to leave one’s institution and ticket to initiate renegotiations of one’s current terms of employment (O’Meara et al., 2017). Many higher education institutions require that faculty first provide an offer letter as evidence of their relative “worth” in their field’s academic labor market if they wish to request a one-time pay raise or other job benefits (O’Meara, 2015a). Thus, faculty rely on outside offers as levers to negotiate resources because once they are hired, a decades-long career on the tenure track might present just two or three scripted moments when it is culturally acceptable to request much more than a cost-of-living adjustment (O’Meara, 2015a). In this sense, outside offers are but one piece of larger faculty reward systems in which pay and resources are scarcely distributed. They represent a “test” of the market for faculty as well: how much another organization is willing to compensate them sets expectations for how far their current institution is (or should be) willing to go to retain them.

Unlike the human resource space, there is very little research on outside offers, who receives them, and the policies that shape and predict them within larger faculty reward systems in higher education (for some examples, see O’Meara, 2015b; O’Meara et al., 2017). Research shows men are more likely to receive outside offers compared to women in general, but rank is the strongest predictor (O’Meara et al., 2017). Full professors followed by associate professors are the most likely to have received outside offers, and men dominate these ranks nationally (O’Meara et al., 2017). And in research universities — where most resources are concentrated in the postsecondary labor market — men sizably outnumber women in these ranks (Snyder et al., 2018).

There are currently no studies of outside offers in higher education through a racial equity lens, though research in faculty hiring and mobility may provide some guidance. Contrary to the canard about being in an enviable market position, studies show that racially minoritized scholars are no more likely to receive lucrative offers or leave early for better opportunities compared to their white counterparts (Smith et al., 1996; White-Lewis et al., 2023). Scholars of color also suffer from narrow conceptions of their research impact (Settles et al., 2020; White-Lewis, 2020), beliefs about their preferences for certain geographic areas (White-Lewis, 2019), and from being less represented among full and associate ranks (Snyder et al., 2018), who by virtue of having longer track records are more likely to have had an outside offer compared to more junior peers (O’Meara et al., 2017). These factors may make it less likely that racially minoritized scholars receive outside offers, and by extension retention offers. Yet, there have also been recent efforts to hire more minoritized scholars through diversity-focused searches (White-Lewis, 2021; Wood, 2021) and strong efforts to retain minoritized scholars as well. If racially minoritized scholars may now be receiving more and/or higher quality outside offers and retention offers, no study to our knowledge provides evidence in either direction.

Research on counteroffers and negotiations

We have established that many universities require faculty to produce an outside offer before they will consider a retention raise (O’Meara, 2015a). While there are no empirical studies on counteroffers or on the equity concerns they raise for higher education, research from human resources and economics sheds some light on similar negotiation processes (e.g., Barron et al., 2006; Capron & Chatain, 2008; Gardner, 2002; Golan, 2005; Landau & Leventhal, 1976; MacKenzie et al., 2012; Postel-Vinay & Robin, 2004; Scott & McMullen, 2017). This research is primarily concerned with the three levels of factors that predict receipt and quality of counteroffers: the current organization, the external organization, and the individuals themselves. Next, we explore how findings in these areas apply to counteroffers in higher education to guide our analyses.

Predicting counteroffers begins with an organization’s stance on whether and how to provide them at all (Barron et al., 2006; Landau & Leventhal, 1976; Postel-Vinay & Robin, 2004; Scott & McMullen, 2017). Few organizations have explicit policies on how to equitably administer them (Scott & McMullen, 2017). As a result, “reactionary and inconsistent decisions can have long-term negative effects on the perceived fairness within the organization and the integrity of its reward program” (p. 7). Some organizations sidestep this by having strict no-counteroffer policies to reduce the likelihood that employees will search (Postel-Vinay & Robin, 2004). For organizations that do provide counteroffers, (1) policies and (2) organizational size have been shown to impact their allocation. For example, Landau & Leventhal, (1976) conducted a simulation of employers’ treatment of employees who presented outside offers. When participants were instructed to follow a policy that emphasized replacing non-productive employees, they gave lower counteroffers compared to others who were told to act on their own discretion in lieu of such a policy. Larger organizations with more employees are also less likely to provide counteroffers, likely because they have other personnel who can perform the duties of outgoing employees (Barron et al., 2006). Although institutional factors are important, the extent to which they predict counteroffers does not appear to outsize other external and individual factors.

Characteristics of the external organization and their corresponding outside offer exert a sizeable influence on who receives counteroffers. The perceived attractiveness of the external organization is one of the chief factors that impacts counteroffers (Gardner, 2002; Landau & Leventhal, 1976; MacKenzie et al., 2012). MacKenzie et al., (2012) conducted a study where participants were either told that the external organization was “one of the more reputable and prestigious firms in our industry…one of the most profitable firms in the industry [or] the firm…filed for bankruptcy protection not that long ago” (p. 385). They found that the size of the counteroffer increased when the outside offer came from the more attractive company (MacKenzie et al., 2012). Even the quality of the offer affects the counteroffer’s reception and quality. Employees who receive more competitive offers, measured by more generous benefits and working conditions, receive better counteroffers (Landau & Leventhal, 1976). These findings make sense, given that pay, benefits, and working conditions are all considered most important in outside offers according to the 2017 Counteroffer Survey completed by personnel in mid- to large-sized organizations outside of higher education (Scott & McMullen, 2017).

Finally, employee-level characteristics partially predict the reception and quality of counteroffers. Studies have considered how counteroffers are shaped by employers’ perceptions of their employees’ productivity and skill (Barron et al., 2006; Landau & Leventhal, 1976), mobility (Gardner, 2002), and position type (Barron et al., 2006; Scott & McMullen, 2017), among other factors. For example, Barron et al., (2006) found that employees who had an above average skillset, higher starting wage, and more years of training were more likely to receive counteroffers. The influence of years of training is particularly interesting to the higher education context because it suggests that organizations that make sizable investments in their employees’ job training and socialization are more reluctant to let them leave. Their study also found that employees who supervised others as part of their job were more likely to get counteroffers. One could criticize that this may co-vary with other significant predictors such as pay and/or training (i.e., supervisors often receive more training, and have higher pay than non-supervisors), but a separate study found that managers across different work settings and pay scales were more likely to receive counteroffers as well (Scott & McMullen, 2017). Though Gardner, (2002) found that perceived mobility also matters, he and other scholars in this area do not nuance this finding by considering how race and gender are constructed in perceptions of mobility in the same way that education and sociology scholars have (O’Meara et al., 2020; Rivera, 2017).

Interdisciplinary research on counteroffers provides numerous hypothetical avenues for how this process might operate in postsecondary careers, but there are shortcomings as well. In terms of organizational characteristics, it may be that larger institutions and departments are less inclined to make counteroffers compared to smaller organizations (Barron et al., 2006). Attractiveness is another standout factor that impacts counteroffers: scholars with outside offers from institutions that are deemed less prestigious may receive lower quality or less expedient counteroffers compared to those with offers from more highly ranked departments and institutions (MacKenzie et al., 2012). Faculty who present more generous outside offers, measured by the percentage increase of pay, may also receive higher quality outside offers as well (Landau & Leventhal, 1976). Regarding individual characteristics, faculty who have been at the institution longer, have higher rank, and have administrative appointments and/or affiliations (i.e., act in a supervisory role) may stand a better chance of receiving better counteroffers (Barron et al., 2006). But the primary limitation for these studies is that they do not factor in race and gender into their analyses. Due to the racialized nature of higher education and faculty careers, it may be the case that some of these hypothesized relationships operate differently for minoritized candidates. Minoritized scholars who receive highly attractive outside offers may still be presumed to be less mobile (Rivera, 2017), conduct work seen as less central (Settles et al., 2020), and are less represented among the ranks (e.g., associate and full professors) and disciplines (e.g., STEM fields with large, supervised labs, centers, and institutes) that “supervise” or lead other faculty and/or students (Snyder et al., 2018).

In this study, we reconcile these gaps by analyzing quantitative data in the Collaborative on Academic Careers in Higher Education (COACHE). In 2015, COACHE launched a new research-practice-partnership with research universities to standardize the data collected and stored about faculty who receive outside offers and identify patterns in counteroffers. The project gathered information about retained faculty and voluntary departures from institutional databases and an online survey instrument whose themes span the search for a new position; the nature of the outside offer; the factors that weigh into a decision to stay or leave; the influence of spouses’ and partners’ careers; the counteroffer process, if any; the transition to a new institution; and inequities in the experiences of faculty at every stage of this phenomenon. Using a subset of these data, we explore the conduct and process of counteroffers and ask critical, equity-driven questions of who receives them and why. Thus, the primary research questions are: (1) Who received a counteroffer, and (2) What are the key predictors of receiving a counteroffer?

Methodology

Data source & approach

Our data source is the COACHE Faculty Retention & Exit Study, a robust and entirely unique source of data for investigating faculty mobility. Faculty must meet several eligibility criteria to be included in the Faculty Retention & Exit Study, and these factors, alongside the specific conditions and questions central to this study, shaped our participant sample. For the purposes of this study “the faculty” refers to the appointment types that were eligible to take the survey, though we have restricted the sample to faculty members who were seeking a counteroffer and either received one or did not. This sample subsequently included full-time, tenure-track assistant, associate, and full professors, or full-time non-tenure-track faculty with multi-year appointments (e.g., those who have voting and senate rights). Other ranks such as endowed professorships were included, though these respondents were identified with their tenure status (i.e., full professor versus endowed full professor) for clarity. It is crucial to note that our analyses do not include part-time faculty, who are the majority of the US professoriate (Eagan et al., 2015).

Overall, 37 institutions participated in the COACHE study from 2016 to 2019 and are reflected in our study. These institutions provided record-level data of the 650 faculty members who sought counteroffers between July 2015 and July 2019. Counteroffer status (i.e., sought a counteroffer and received one or sought a counteroffer and did not receive one) was the primary outcome variable for all analyses and we focus on this subset of faculty respondents to understand what factors may have contributed to the differential counteroffer decisions.

The analytic sample is comprised of 650 faculty members who sought a counteroffer and either received one or did not. This sample is described below in Table 1.

Table 1 Analytic sample demographics (n = 650)
Full size table

Given the relatively low numbers of specific racial groups, we created a faculty of color variable, and with that the dataset was 67.6% white and 29% faculty of color.

While the Faculty Retention & Exit Survey instrument collects quantitative survey responses and qualitative survey comments, we focused specifically on the quantitative variables collected with respect to counteroffers to explicitly respond to our research questions. Our analyses included descriptive analysis of the faculty members who received counteroffers versus those who did not, a predictive regression model describing the predictors that differentiated counteroffer decisions, and a descriptive analysis of the success of counteroffers. We describe the analytic approaches in greater detail below.

Logistic regression

Prior to conducting the logistic regression models, we used descriptive statistics to assess whether there were significant differences between groups in receiving a counteroffer from their institution. After the analyses of variance, we employed logistic regression on the sample (n = 650) to identify whether any variables significantly predicted our primary dependent variable counteroffer receipt, measured by either receiving one or not. Our variable selection was guided by the literature. We included gender (coded as women and men—while there were nonbinary respondents in the sample, they did not respond to the counteroffer question), race (coded as faculty of color or white, given the lack of significant difference among the groups in our analyses of variance), tenure status (coded as untenured versus tenured), institutional type (R1 versus R2 and R3), academic area (coded as dummy variables for each individual academic area: humanities, social sciences, STEM, and professions and other fields), career age measured as years since the completion of the respondents’ terminal degree, marital status, whether the faculty member communicated with a member of their department (i.e., a chair, dean, or other leader) regarding their external offer (coded as contacted or did not contact), the respondent’s salary at the time of their external offer, the offered salary at the external institution, salary change as a percentage difference (e.g., were they getting a 50% raise to leave their institution?), whether there was a startup package included in their external offer (0/1), and how seriously they were considering leaving prior to receiving a counteroffer (or not). We entered all variables together in one block against the dependent variable, counteroffer receipt.

Model fit

To determine the fit of our model, we employed Akaike information criterion and calculated a Pseudo R-squared. The best-fit model (which included all variables listed above) carried 90% of the cumulative model weights, the residual deviance was significantly lower than the null deviance, and the final model had a Pseudo R-squared of 0.35. While this explained variance is relatively low, the overall model still holds explanatory power. The results of both approaches are presented below.

Counteroffer receipt and success

While the goal of this paper was to clearly describe who receives counteroffers and identify key predictors of receiving a counteroffer, we also considered the success of counteroffers in retaining faculty members. Of the 650 faculty members in the dataset who solicited a counteroffer, 331 failed to receive one from their institution. 325 (98%) of these faculty departed their institutions while 6 (2%) remained. Of the 319 faculty members who received a counteroffer from their institution, 135 (42%) departed their institutions while 184 (58%) were successfully retained. Due to the early nature of our research on counteroffers and their effects, describing the circumstances under which these counteroffers were successful is outside the scope of the current paper; as a result, our work has some limitations and leaves aside some unanswered questions for future research. We describe these limitations below and discuss future research within the discussion and conclusion of this paper.

Limitations

There are inherent limitations to this study. First, similar to other studies using secondary data, we were limited by the variables and responses in the COACHE Faculty Retention & Exit Study dataset. While some variables would have been ideal to test in our logistic regression, such as measures of institutional and department prestige for the external and current institutions in the counteroffer process (e.g., Matier, 1990; O’Meara et al., 2014), there were no variables describing this. The most obvious choice to capture institutional and departmental prestige would have been to impute data from popular ranking systems such as U.S. News and World Reports, but a few factors limited this decision. First, the recent destabilization of institutional ranking regimes in the United States means that fewer institutions are participating (Diep, 2022). This, coupled with the fact that departmental-level ranking data are few and far between, making the creation of a proxy for departmental prestige exceedingly challenging and subjective, led us to omit prestige as a potential variable to limit bias in the model.

Additionally, there were some variables within the dataset that were interesting, such as the dollar amount and number of grants a faculty member was awarded, but were not included in the model due to data missingness. COACHE invites universities to attach institutional data about individuals’ grants, research productivity, teaching load, and departmental prestige. These factors could, in neoliberal terms, describe the “attractiveness” of the faculty member in the academic labor market. These institutionally collected and reported variables, however, are particularly prone to missingness and have not yet been provided to the authors in a state suitable for analysis. This study also does not consider whether certain qualities of the outside offer (e.g., perceived prestige of originating institution) or the ability of current institution to afford a counteroffer affect whether or not a counteroffer is presented at all. Several variables in the COACHE dataset do describe qualities of the outside offer and counteroffer; we include salary in this study, but other qualities of the outside offer and counteroffer (e.g., summer salary, amount of additional startup, course release, spousal/partner employment) are also theoretically available. Given the scale and complexity of these factors, particularly the variability with which these metrics are reported by participating institutions, we are conducting a parallel line of inquiry that extends these findings.

Our analyses are also limited by the number of faculty who solicited and received a counteroffer (16%) and thus are likely limited by various types of nonresponse bias (Rogelberg et al., 2003). While the analytic sample is comparable to the overall dataset, demographic data were to some extent missing for independent variables like gender (12.8%) and tenure status (21.7%), though these variables were supplemented by data collected in the survey. Because there is no discernible pattern or rationale to explain why institutions failed to provide certain data (institutions were bound by resource constraints, legal statutes restricting the use of employees’ demographic data, logistical difficulties, etc.), the demographic data missing appears to have been largely missing at random.

The demographic descriptive statistics described in the “Methodology” section, however, suggest that the analytic sample focusing on counteroffers is comparable to the larger sample and that women and white faculty may be somewhat overrepresented in the respondent groups. Any potential effects of nonresponse bias that might have resulted in data skew are potentially active nonresponse—the explicit choice by subjects not to participate in the survey—which could be associated with fears regarding potential retaliation. However, we expect active nonresponse to be a small fraction of total nonresponse (Mathews, 2013; Rogelberg & Stanton, 2007). Finally, the percentage of faculty who sought counteroffers was a significantly smaller subset of the total sample, though the number of those who received counteroffers and those who did not was nearly equal. While it would have been preferable to have a larger sample, the analytic sample (n = 650) was more than adequate for our purposes.

Findings

Descriptive statistics

To answer our first research question, we turned to descriptive statistics to analyze whether there were significant differences between groups when it came to receiving a counteroffer from their institution. While the descriptive statistics presented earlier in this section indicate that the analytic sample was predominantly white, more men than women, more tenured than not, and from high research activity institutions, frequency distributions cannot indicate whether there were significant differences between these groups. We present the results of these descriptive statistics below as they influence the subsequent analyses (logistic regression).

As we were interested in differences between particular groups (i.e., women and men faculty, tenured and untenured faculty, faculty of color and white faculty, etc.), we conducted one-way analysis of variance (ANOVA) tests to determine if there were significant differences among these groups. Our first ANOVA considered counteroffer receipt and different race groups. There was no significant difference between the groups when disaggregated. Our second ANOVA considered two pools: white and faculty of color. There was a statistically significant difference among the two groups on counteroffer receipt (F = 3.452, p < 0.05), with fewer counteroffers made to faculty of color. Our third ANOVA considered counteroffer receipt and faculty at different institutional types. There was a statistically significant difference among the two groups on counteroffer receipt (F = 6.369, p < 0.01), with faculty at less research intense institutions receiving fewer counteroffers. Our fourth ANOVA considered counteroffer receipt and gender. There was no significant difference between men and women faculty with respect to counteroffer receipt. Our fifth and final ANOVA considered counteroffer receipt and tenure status. There was a statistically significant difference among the two groups (tenured and untenured) with respect to counteroffer receipt (F = 20.74, p < 0.001). Untenured faculty received significantly fewer counteroffers than tenured faculty members.

Logistic regression

The results of the logistic regression are presented in Table 2. Initial coefficients, as well as Odds Ratios, are presented. Due to limitations in the data, race was coded as either being white or faculty of color, while gender was coded as either being a man or a woman. Tenure status was coded as untenured or tenured. All other variables were continuous or binary. Finally, we considered interaction terms based on the results of the analyses of variance, though the interaction terms for untenured racially minoritized faculty and racially minoritized women faculty did not significantly improve the model, so they were removed from the final results.

Table 2 Logistic regression: counteroffer reception
Full size table

The logistic regression focused on determining whether there were any predictors of counteroffer receipt among our sample of faculty who solicited counteroffers (n = 650, pseudo-R-squared = 0.35). We found several statistically significant relationships.

We began by piecing together and control for the beginning of the counteroffer process, by using the variable “informed department of external offer.” Faculty members who informed their department or school of their external offer were much more likely (40 times) to receive a counteroffer than those who did not (p < 0.0001). This may seem obvious, but given that other variables were still statistically significant after accounting for a variable with such high explanatory power, clearly there are other factors that shape the conduct and process of receiving counteroffers other than simply notifying leadership. For instance, there was a statistically significant relationship (p < 0.01) between counteroffer receipt and race. White faculty members were 64% more likely to receive counteroffers than their faculty of color peers. Regarding other personal identity factors, men were nearly 80% more likely than women to receive counteroffers (p < 0.001), and tenured faculty were more than 3 times as likely to receive counteroffers than untenured faculty (p < 0.0001).

While discipline appeared to have a significant effect on counteroffer receipt, the positive relationship between counteroffer reception and the three disciplines that were statistically significant and the fourth that was not muddies whether a particular field engages in the conduct of counteroffers more than another. Specifically, faculty members in the humanities were more likely to receive counteroffers than their colleagues in the social sciences, STEM fields, and professions. Because each variable was coded as binary (i.e., a humanities faculty member or not), the statistically significant results for three of the four included fields stand in contrast to each other and are generally inconclusive. Interactions between field, race, and gender, were similarly inconclusive and omitted from the model due to lack of statistical significance, suggesting that field may be a confounding variable when it comes to receiving a counteroffer.

While results for field were inconclusive, career age—or the time between a faculty member’s receipt of their terminal degree and their current career stage—negatively predicted receipt of a counteroffer, with older faculty being marginally less likely (8%) to receive a counteroffer for each year of their career (p < 0.0001). With respect to institutional type, faculty at less research-intense universities were 80% less likely to receive a counteroffer (p < 0.001). Finally, faculty whose external offer came with a startup package were 74% more likely to receive a counteroffer (p < 0.01) and faculty who genuinely wanted to leave their institutions were significantly less likely—30%—to get a counteroffer (p < 0.0001). We discuss these results below.

Discussion

Responding to public calls for greater institutional transparency, and attention to equity issues in higher education, many universities are re-examining how they hire, retain, and reward faculty. This is evident in spaces such as promotion and tenure (O’Meara et al., 2022), hiring (Liera & Hernandez, 2021; Culpepper et al., 2023), and workloads (O’Meara et al., 2022). But one of the least examined areas of faculty affairs — either from research, efficiency, or equity perspectives—— are outside offers and counteroffers. Acker (1990) advocated that those interested in advancing equity in organizations “laser in” on precise spaces and mechanisms that are reproducing inequality. Counteroffers provided to faculty at research universities are important to study because they can either be conducted (a) effectively and equitably, signaling to faculty that they are valued in the hopes of retaining valuable talent for the organization, or (b) poorly, resulting in a host of negative outcomes such as losing talent, retaining talent at too high a cost, reducing organizational legitimacy, and/or inviting potential legal discrimination suits.

The organizational backdrop within which counteroffers are conferred are discretionary spaces, making either form of conduct a realistic possibility. Lipsky (1980, 2010) first introduced the concept of discretionary spaces in his book Street Level Bureaucrats. Lipsky observed that professionals are given discretion to make decisions in contexts where expertise, flexibility, and judgment are needed, and where automated responses will not be effective or equitable. For example, it would not be effective for a department chair to have the same response to each and every faculty member who came to them with an outside offer, because each faculty member will have different levels of productivity and will be separated by different subfield norms with varying labor market constraints, requiring individualized decision-making. Thus, department chairs and deans are given discretion to take these different contexts into account when making decisions about when and how much to provide in a counteroffer.

However, decision-makers can make discriminatory judgments — especially without oversight — and at minimum carry cognitive and social biases that allow discretion to be enacted in ways that disadvantage some groups and advantage others. Ball, (2018) applied this idea to math classrooms, showing how teachers could use their discretion to reinforce racial stereotypes, or combat them in equity-minded ways. O’Meara, (2021) applied this concept to higher education, showing how the discretion of academic leaders and faculty might be leveraged, checked, and/or restructured to create more equitable organizations. She and others find that there are certain conditions that can be installed to safeguard equity, yet today none of these (e.g., transparency, consistency, accountability) are documented in counteroffers. Instead, counteroffers are made with all the trappings shown to reproduce inequality. Using data from the COACHE Faculty Retention & Exit Study, we focused on several components of this under-examined space. Although there were many findings that illuminate the process of counteroffers and the characteristics of who receives them, we unpack the four most prominent findings.

Given the paucity of research on how identity characteristics impacts the conferral of a counteroffer, a noteworthy contribution of this study was revealing early inroads into how these factors matter. We found that white faculty received more counteroffers than faculty of color and were 64% more likely to receive them. We also found that male faculty members were 80% more likely than women faculty members to receive counteroffers. Given that all of these faculty were competitive enough to receive outside offers, these results are concerning from an equity perspective. Although addressing the mechanism that drives this disparity is outside the scope of the current study, extant research suggests that the cause is multifactorial. That is, research shows that minoritized faculty are already disadvantaged during negotiations (Hernandez et al., 2019; Toosi et al., 2018), are perceived as less mobile (Rivera, 2017; O’Meara, 2021), and occupy fewer positions seen as valuable to the university, such as senior leadership (Sagaria, 2002), and senior faculty positions (Snyder et al., 2018). The latter explanation is supported by our finding that tenured faculty were three times as likely to receive counteroffers compared to their untenured peers. This makes sense from an organizational perspective, since institutions make sizable investments in their tenured faculty, and these investments may also include counteroffers.

Related to organizational factors, the faculty member’s current institution in many respects sets the terms for engagement related to individual counteroffer conduct. It is generally understood nowadays that institutions have a counteroffer mechanism in place, so we were less interested in whether the presence of a formal policy predicted counteroffers (e.g., Landau & Leventhal, 1976; Scott & McMullen, 2017). Instead, we found that faculty who notified their leadership of an external offer were significantly more likely to receive a counteroffer than those who did not. On the surface this may seem obvious: deans and department chairs cannot make retention offers if they are unaware that there is a clear and present risk of their faculty member resigning. However, this can also speak to whether faculty perceive there to be a departmental and/or institutional climate wherein they feel comfortable enough to share that they received an external offer. This is even more pronounced in many small fields and disciplines where academic leaders may hear of a faculty members’ potential exit through sources other than said faculty member. We also found that faculty at less research-intensive universities were 80% less likely to receive a counteroffer. This likely correlates with the fact that research universities often have greater resources to retain their faculty compared to less research-intensive institutions, which aligns with the literature on how larger organizations with more resources are more likely to provide counteroffers (Barron et al., 2006).

Equally as important if not more so than the faculty member’s current institution is the organization making the external offer (Gardner, 2002; MacKenzie et al., 2012). We found somewhat conflicting results regarding the quality of external offers in predicting a counteroffer.

Contrary to prior literature (e.g., Landau & Leventhal, 1976), we did not find that the size of the salary increase proposed by the external organization had any influence on whether a counteroffer materialized from the home institution. There may be some reasons for this. On the one hand, a proposed salary increase may not influence whether or not a counteroffer emerges due to extraneous factors such as cost-of-living. An outside offer from a state where the cost-of-living is much higher may not necessarily motivate institutional leaders to work any harder than they would have already; an outside offer already represents a signal that the faculty member has the means to leave, and the amount may not necessarily add any more value than the offer already represents. And at worse, a very high proposed salary may discourage an institution from attempting to retain them, especially if there was an assumption that it would be impossible to provide similar compensation. In this sense, a substantial outside offer may operate as a deterrent, though more research is needed to substantiate this possibility.

However, we did find that faculty whose outside offer came with a startup package were 74% more likely to receive a counteroffer than those whose offers did not. Taken together, there seems to be inconclusive evidence on how strength of the external offer impacts counteroffer reception. We believe this finding can be explained by important distinctions between salary and startup funds in resource-constrained environments for two reasons. First, most universities, especially public institutions, must follow guidelines for faculty salary scales that discourage, if not outright limit, “off-scale,” or “out-of-step” pay raises for faculty of comparable status. But these guidelines do not apply to startup fund allocations, which do not appear in pay equity reviews or salary disclosures. Second, providing startup funds to an individual faculty member is a one-time outlay from a budget, whereas salary increases — and the basis it creates for fringe benefit assessments — compound year over year until the professor retires. A professor’s final salary before retirement may determine pension or other retirement benefits, too, making each retention decision a costly one that a university may endure for the rest of the faculty member’s life. Considered this way, an outside offer with a startup package might be more readily countered by the home institution with its own increase in startup benefits, benefits from which the entire community may also benefit (e.g., graduate student support, improved facilities and equipment), which are not typically realized by salary increases that benefit of private individuals. This point is buoyed by literature that shows that a faculty member’s decision to seek an external offer is not always driven by salary, but other factors (White-Lewis et al., 2022).

Implications for future research and practice

Based on the results of our study, we see several areas ripe for future research. Our study is among the first to significantly document how identity characteristics such as gender and race matter in receiving a counteroffer. Future research should examine possible moderating variables such as grant productivity, family status, perceived mobility, and perceived relationships with colleagues in the department. It may be that some of these factors are driving our results, or not. For example, if grant productivity was unrelated to our findings showing that minoritized faculty received fewer counteroffers (e.g., minoritized faculty receiving fewer counteroffers despite notable grant records) than that would make our study’s findings even more pressing. Another type of quantitative study could utilize experimental vignette methods to test how different configurations of departmental policies drive counters. Similar to how Landau & Leventhal, (1976) manipulated a counteroffer policy, there may be room to investigate how departmental and institutional policies can create guardrails to ensure consistency and equity.

We also see high value in conducting qualitative research in this area. Ethnographic and/or interview-based studies of department chairs and deans responsible for making such decisions are necessary to demystify the conduct of counteroffers. These studies would ideally reveal if there are consistent practices and routines across leaders within a single institution or across multiple institutions and institutional types. There may be strong repertoires of practice (Posselt et al., 2020) that guide decision-making, or it could be an even wider discretionary space than we initially imagined. For instance, the timing in which leaders act on outside offers may vary, or how leaders weigh and consider the relative prestige of the outside offer and use those factors to discern the risk of the faculty member leaving or not. Given the evidence in the present study and extant literature on inequities in academic departments, it is important to uncover drivers that may create different counteroffer processes based on gender, race, and tenure status.

Concerning practice, we are inspired by repertoires of practice (Posselt et al., 2020), which underscore the critical need to create more opportunities for equity within organizational routines. In making constructive, actionable implications for practice we face the same conundrum that currently animates the faculty hiring discourse: though there are clearly documented racial and gender inequities in hiring criteria like grantsmanship (e.g., Chen et al., 2022), very few search committees, if any, are willing to hire without considering such factors. Thus, current work in that space advocates for manipulating the items, weight, and timing of those assessments in order to introduce more equitable evaluations (Culpepper et al., 2023). We see a similar problem in this space; that is, overly restrictive standardization is unlikely to happen as it would compromise the flexibility that chairs and deans need to deliver counteroffers quickly, but the results of our study underscore how this flexibility compromises equity. Though some advocate for policies in which nobody receives a counteroffer (similar to research on strict no-counteroffer policies; Postel-Vinay & Robin, 2004), we see issues with this approach as well.

One novel solution is that every faculty member who presents an outside offer is assigned an advocate who will work with them to strike the best deal possible. The advocate would be an expert in negotiations and the context in which the university operates, so they will be able to guide them on what is possible within that context. Under the current system, institutional leadership sees every potential departure — increasing their negotiating expertise — whereas the potentially departing faculty member only experiences this phenomenon once, maybe twice in their career. This would help “even the playing field” so that the faculty member is also equipped with the institutional history, backdrop, and knowledge to more effectively negotiate. Another beneficial step would be to improve record-keeping standards for outside offers and counteroffers to help institutional leaders — and negotiation advocates — to discern trends among those whom they were able to retain and those they could not. Better records could include the speed with which the counteroffer was made, the subfield, and the quality of interactions (including and beyond negotiations) with the faculty member. These data would ideally reveal areas of success and opportunities for improvement to drive equitable and effective praxis.

At minimum, institutional leaders should improve onboarding practices for department chairs and deans to prepare them for the counteroffer process. Any sort of preparation materials or training should require that the process be conducted consistently (e.g., using the same data for each decision), clarity (e.g., transparency around the timeline in which the decision should be made, who has authority to approve the decision, the amount and various forces that shape the amount either positively or negatively such as external market forces or institutional budget constraints). These are important tenets that ensure that the process is fair. But to be more equitable, institutions must go a step further to collect and leverage data to expose inequities that impact the careers of minoritized scholars, and create cycles of inquiry (e.g., Dowd & Liera, 2018) that use those data to improve the counteroffer process. Overall improvements to the system are needed, but particular attention must also be given to those impacted most. If cycles of inquiry reveal persistent disparities by race, gender, or any other factor, academic and administrative leaders must use those data to develop and/or widen their available toolset to directly address minoritized colleagues most impacted.