Alert Top Message

Due to concerns about COVID-19, the AAUP office has transitioned to telework. Please contact staff by email.



Policies on Academic Freedom, Dismissal for Cause, Financial Exigency, and Program Discontinuance

 I. Introduction

A central goal of the AAUP is to protect academic freedom, tenure, and due process by assisting faculty governance bodies and AAUP chapters in their efforts to incorporate AAUP-recommended policies in faculty handbooks and collective bargaining agreements. The AAUP achieves this goal in various ways, from directly assisting chapters and other faculty bodies in devel­oping contract and handbook language to providing guidance on interpreting these policies. The enforce­ment of Association-recommended policies through the mechanism of investigation and censure also plays a role in the adoption of such policies.1

This report provides a statistical analysis of the presence of AAUP-recommended policies on academic freedom, dismissal for cause, financial exigency, and program discontinuance in faculty handbooks and collective bargaining agreements. In the best of times, analysis of their prevalence could usefully inform the work of AAUP chapters, faculty governance bodies, and higher education unions, but given the impact of the COVID-19 crisis on campuses around the coun­try, these data are now even more important to the advancement of the AAUP’s principles and policies. Statistical evidence of the widespread adoption of AAUP policy statements in faculty handbooks and contracts can reinforce the argument that institutional practices that depart from AAUP-supported standards are outside of the mainstream. Conversely, informa­tion about which institutional policies more frequently fall short of Association-recommended policies can be useful for faculty members engaged in reviewing regulations or contracts.

In 2000, Cathy Trower, who at the time was a researcher at the Harvard Graduate School of Education, edited a book in which she and a group of collaborators presented results of a survey of faculty appointment policies in faculty handbooks and col­lective bargaining agreements based on a stratified random sample of 217 four-year institutions of higher education.2 The population from which the sample was drawn consisted of four-year institutions classified as bachelor’s, master’s, doctoral, and research institu­tions in the then most recent Carnegie classification system. The study compared institutional policies to applicable AAUP standards on academic freedom, tenure, and due process and reported on the preva­lence of various types of policies.

The analysis conducted for this report partially rep­licates Trower’s study to provide updated information about the prevalence of several appointment-related policies and to track changes that have occurred during the past two decades. Differences in preva­lence based on Carnegie classification and on whether the faculty at the institution engages in collective bargaining provide important context for such find­ings and will be reported where relevant. The policy areas considered in this report are

  • the provenance of academic freedom state­ments,
  • grounds for dismissal for cause, and
  • policies related to terminations of appointment because of financial exigency and program discontinuance.

Because of the close relationship of these types of policies to tenure, Trower restricted her analysis to institutions that had tenure systems. Of 217 institu­tions in her sample, 196 had a tenure system. This report has also restricted the analysis to institutions with tenure systems for the same reason and to make current results comparable to the prior findings. The analyses in this report are based on a sample of 198 institutions with a tenure system; 174 of those institutions (89 percent) were also in Trower’s sample. Details about the sample and other methodological considerations can be found in the appendix.

It is important to note that both faculty hand­books and collective bargaining agreements can have varying degrees of legal enforceability. Although their legal status is not uniform, in some jurisdictions faculty handbooks are binding contracts, enforceable in court.3 Provisions in collective bargaining agree­ments are generally enforceable through a final and binding arbitration process, but this is not always the case. It is also worth bearing in mind that the scope of bargaining has different limitations in the vari­ous statutes that enable private- and public-sector collective bargaining. Thus, provisions addressing certain topics discussed here may be missing from a negotiated agreement specifically because they fall outside of the scope of bargaining. I made efforts to find applicable policies in other institutional regula­tions when necessary, and, in the case of academic freedom statements, I separately tracked whether the statement was located in the collective bargaining agreement or in other regulations, since the state­ments’ location may affect enforceability. However, in its assessment of institutional regulations rela­tive to Association-supported standards, this report considers neither differences in enforceability nor limitations to the scope of bargaining, although those factors certainly matter to the overall assessment of such regulations. At some institutions, the regulations considered here apply to faculty members serving on contingent appointments, and at others there are separate regulations for those faculty members or even none at all. This study did not assess whether the policies analyzed here apply to faculty members on contingent appointments.

The Prevalence of Tenure and the Composi­tion of the Population

According to the Integrated Postsecondary Education Data System (IPEDS), in 2018 the United States had 1,308 four-year public or private not-for-profit insti­tutions of higher education classified as bachelor’s, master’s, or research/doctoral institutions. Excluded from this population are two-year colleges, for-profit institutions, and specialized institutions, such as semi­naries or free-standing law schools, some of which have tenure systems as well. According to IPEDS, 89 percent of the institutions in this population, a total of 1,170, report having a tenure system. The prevalence of tenure differs by institutional type: it is essentially universal at research institutions, and it is highly prevalent at both master’s and bachelor’s institutions (see figure 1). Again, following Trower’s study, I have designed the analyses in this report to be generalizable only to the 1,170 four-year institutions that have a tenure system.

The sample makes it possible to estimate the preva­lence of collective bargaining overall and according to institutional type (figure 2).4 Overall, tenured and tenure-track faculties at 19 percent of institutions that have a tenure system engage in collective bargaining. Collective bargaining is much more common among master’s institutions (29 percent) than among bach­elor’s institutions (8 percent), with the prevalence at research institutions in between (18 percent). The relative rarity of collective bargaining at bachelor’s institutions is, of course, related to the high prevalence of private control among those institutions (84 percent of bachelor’s institutions are private, compared with 34 percent of research institutions and 37 percent of master’s institutions), given that collective bargaining has been exceedingly uncommon at private institu­tions since the Supreme Court’s 1980 decision NLRB v. Yeshiva University, in which most full-time faculty members in private institutions were denied the right to pursue collective bargaining under the legal frame­work of the National Labor Relations Act.

Academic Freedom

Throughout US higher education, the 1940 State­ment of Principles on Academic Freedom and Tenure, formulated jointly by the AAUP and the Association of American Colleges and Universities and endorsed by more than 250 disciplinary societies and educa­tional associations, serves as the locus classicus of the definition of academic freedom. The 1940 Statement contains the following three provisions on academic freedom:

  1. Teachers are entitled to full freedom in research and in the publication of the results, subject to the adequate performance of their other aca­demic duties; but research for pecuniary return should be based upon an understanding with the authorities of the institution.
  2. Teachers are entitled to freedom in the classroom in discussing their subject, but they should be careful not to introduce into their teaching con­troversial matter which has no relation to their subject. Limitations of academic freedom because of religious or other aims of the institution should be clearly stated in writing at the time of the appointment.
  3. College and university teachers are citizens, mem­bers of a learned profession, and officers of an educational institution. When they speak or write as citizens, they should be free from institutional censorship or discipline, but their special posi­tion in the community imposes special obliga­tions. As scholars and educational officers, they should remember that the public may judge their profession and their institution by their utter­ances. Hence they should at all times be accurate, should exercise appropriate restraint, should show respect for the opinions of others, and should make every effort to indicate that they are not speaking for the institution.

The AAUP has supplemented this definition of aca­demic freedom in many subsequent policy statements, including the 1970 interpretive comments, which the AAUP publishes together with the 1940 Statement, and interpretations in AAUP investigative reports that deal with violations of academic freedom and ten­ure. The 1940 Statement thus constitutes the central element of the AAUP’s policy on academic freedom and tenure, but it is still only part of a larger body of related policies. It is desirable, from the perspective of the Association, that academic freedom provisions in faculty handbooks and contracts be interpreted in light of the entirety of AAUP policies.

With respect to academic freedom statements, Trower’s study categorized institutional regulations into four groups:

  1. those that explicitly cite the 1940 Statement or quote extensively from it with attribution to the statement or to the AAUP
  2. those that quote extensively from the statement without attribution
  3. those that do not use language from the 1940 Statement
  4. those that do not include a statement on academic freedom

In addition to providing a taxonomy of academic freedom statements, the four categories can arguably be regarded as forming a hierarchy with respect to adherence to AAUP standards. As noted above, the AAUP’s policies on academic freedom and tenure take the 1940 Statement as their point of departure. Direct inclusion of the 1940 Statement and attribution to its AAUP source facilitate the argument that exist­ing academic freedom language in faculty handbooks or collective bargaining agreements should be inter­preted in light of derivative AAUP policy statements or investigative reports. Quoting the 1940 Statement directly, even without attribution, also facilitates such an argument.

This study found that the 1940 Statement con­tinues to serve as the primary source for academic freedom language in institutional regulations (figure 3): not only do almost three-quarters of institutions with a tenure system (73 percent) base their academic freedom policy directly on the 1940 Statement, but more than half cite the AAUP as the source of their policy. Moreover, as figure 4 indicates, the prevalence of academic freedom policies attributed to the 1940 Statement has increased from 45 percent to 52 percent compared with Trower’s study of 2000, while the number of institutions without an academic freedom statement has decreased from 8 percent to 3 percent. In both the 2000 and 2020 studies, 24 percent of institutions have academic freedom state­ments not based on the 1940 Statement. In light of the hierarchical view of the categories proposed above, the overall findings indicate positive develop­ments, especially the overall finding regarding the impact that the AAUP has had on academic freedom language: one would be hard-pressed to identify any other language contained in three quarters of all faculty handbooks and contracts.

 The prevalence of academic freedom statements varies both by institutional type and by the faculty’s collective bargaining status. Research institutions more frequently use academic freedom statements not based on the 1940 Statement (37 percent). Among bachelor’s institutions, 85 percent have academic freedom statements based on the 1940 Statement, with 61 percent attributing the statement to its source. Fifty percent of master’s institutions and 42 percent of research institutions attribute their academic freedom statements to the AAUP source.

It is worth noting that all of the collective bargaining institutions have academic freedom state­ments of some kind, while 4 percent of institutions without a faculty union lack academic freedom state­ments. On the other hand, the inclusion of statements not based on the 1940 Statement is more common at institutions with faculty unions than at those with­out. Additionally, 79 percent of institutions in which the faculty engage in collective bargaining incor­porate the academic freedom statement into their contracts.

Dismissal for Cause

The AAUP has long held that protecting academic freedom requires faculty handbooks and collective bargaining agreements to include specific safeguards against arbitrary dismissal of faculty members. Poli­cies governing faculty dismissals consist of procedural elements, such as those relating to the selection and composition of the faculty hearing body, and substan­tive elements—in particular, what qualifies as a ground for dismissal. The analysis in Trower’s volume focused only on the substantive grounds for dismissal, and my analysis will thus proceed in the same way.

The Statement on Procedural Standards in Faculty Dismissal Proceedings, also jointly formulated by the AAUP and the Association of American Colleges and Universities, is the AAUP’s primary policy statement on faculty dismissal. It makes the following observa­tion about grounds for dismissal:

One persistent source of difficulty is the defini­tion of adequate cause for the dismissal of a faculty member. Despite the 1940 Statement of Principles on Academic Freedom and Tenure and subsequent attempts to build upon it, consider­able ambiguity and misunderstanding persist throughout higher education, especially in the respective conceptions of governing boards, administrative officers, and faculties concerning this matter. The present statement assumes that individual institutions will have formulated their own definitions of adequate cause for dismissal, bearing in mind the 1940 Statement and stan­dards that have developed in the experience of academic institutions.

As the above quotation notes, the 1940 Statement provides little guidance on acceptable grounds for dismissal. In fact, it lacks a full enumeration of such grounds, naming only “incompetence” and “moral turpitude.” Subsequent AAUP policy documents and investigative reports have focused on grounds for dis­missal that the Association has deemed unacceptable, including, for example, insubordination, membership in a political party (such as the Communist Party), and refusal to swear a disclaimer or loyalty oath.

The derivative Recommended Institutional Regulations on Academic Freedom and Tenure, which contains formulations of the AAUP’s procedural standards in a form suitable for direct incorporation into faculty handbooks and collective bargaining agreements, provides the following language on grounds for dismissal: “Adequate cause for a dismissal will be related, directly and substantially, to the fitness of faculty members in their professional capacities as teachers or researchers. Dismissal will not be used to restrain faculty members in their exercise of academic freedom or other rights of American citizens.” Although not providing a definition, the above language places limitations on what institutions can employ as adequate grounds for dismissal, and some institutions quote this provision in their regulations even if they go on to define “adequate cause” in further detail.

One additional source of policy language on dismissal is the 1973 report of the joint Commission on Academic Tenure in Higher Education, which was sponsored by the AAUP and the Association of American Colleges and Universities.5 The commission recommended that grounds for dismissal be restricted to “(a) demonstrated incompetence or dishonesty in teaching or research, (b) substantial and manifest neglect of duty, and (c) personal conduct which sub­stantially impairs the individual’s fulfillment of his [or her] institutional responsibilities.”

This study shows that both the Association’s position on grounds for dismissal and the recom­mendations of the commission are reflected in a large percentage of institutional regulations, some of which use the commission’s formulation verbatim, although never with attribution.

Of the four grounds for dismissal listed by the commission, incompetence, neglect of duty, and dishonesty (together with falsification and misrep­resentation) are the first, second, and fourth most common across all institutions (see figure 5): 65 per­cent of institutional regulations list incompetence; 57 percent list neglect of duty; 34 percent list dishonesty, falsification, or misrepresentation; and 19 percent list personal conduct. References to personal conduct (or “personal misconduct”) in institutional regulations at times do not qualify conduct that “substantially impairs” the faculty member’s “fulfillment of his [or her] institutional responsibilities,” which the joint commission’s language included. With or without that qualification, grounds for dismissal related to alleged personal misconduct are the ninth most frequent across all institutions.

Figure 5 presents the overall findings concerning the prevalence of grounds for dismissal. Significantly, the grounds for dismissal that the AAUP has gener­ally viewed as acceptable are the most common. Only the four least prevalent (no definition of “adequate cause,” performance-related, insubordination, and inefficiency) raise concerns relative to AAUP-supported standards, with the lowest-ranking two raising the most significant concerns. Inefficiency was found in only 3 percent of regulations, and other grounds not included in the figure are at odds with AAUP-supported standards are even less common. It is important to clarify that “no definition of adequate cause” has a special role in this analysis: even though institutions can (and usually do) have more than one of the grounds for dismissal listed in figure 5, those that do not define “adequate cause” cannot include any of the other listed grounds in their regulations. In other words, institutions are included multiple times in the categories above, except for the case of those that give “no definition of adequate cause.”

A few observations concerning grounds for dis­missal follow.

No definition of “adequate cause.” The prevalence of institutions that do not provide a definition for “adequate cause” in their official policies has increased from 11 percent to 18 percent in the twenty years since Trower’s study (figure 6). Because the Association has not articulated a complete definition of “adequate cause,” if institutional regulations leave the term undefined, AAUP-supported standards do not provide a complete definition upon which to rely (contrary to the situation for financial exigency, which will be discussed below). A majority (57 percent) of institutions with unionized faculties have institutional policies that do not define “adequate cause.” The prevalence of not having a definition of “adequate cause” varies among institutions without collective bargaining based on insti­tutional type. Bachelor’s institutions without collective bargaining almost universally define “adequate cause” (3 percent do not define it), and 12 percent of research institutions and 14 percent of master’s institutions without collective bargaining do not define it. Contrary to the situation outside of collective bargaining, union contracts that specify only that dismissals can occur for “adequate” or “good” cause generally have arbitral standards to rely on, which likely explains the differ­ence in prevalence.

Certainly, the absence of either arbitral standards or the general limitation on grounds for dismissal in the Recommended Institutional Regulations raises concerns that dismissal policies that lack a definition of “adequate cause” may not protect academic free­dom sufficiently.

Violation of institutional policies or the rights of oth­ers. This category saw an increase in prevalence from 22 percent in 2000 to 30 percent in 2020 across all institutions (figure 6). This change may reflect an increase in institutional policies specified outside of the dismissal policy that may serve as grounds for dismissal, although this study did not separately assess that possibility. As Trower’s study noted, at some institutions, the dismissal policy explicitly cites sexual harassment as a ground for dismissal policies, while at other institutions, dismissals for sexual harassment are treated separately. For that reason, Trower’s study did not include violations of sexual harassment policies in the analysis of grounds for dismissal, and neither does this report. The present category includes only generic statements to the effect that violations of (unnamed) institutional policies or violations of the rights of oth­ers are grounds for dismissal. Examples of the latter encountered in the analysis include the following:

  • “deliberate and serious violation of the rights and freedom of fellow faculty members, admin­istrators, or students”
  • “conduct that interferes with the rights and privileges of another member of the college community”
  • “knowing or reckless violation of the rights and freedom of students or other employees of the university”

Moral turpitude. The frequency of this term in dismissal policies has declined since 2000 from 33 percent to 25 percent (figure 6), which may reflect a sense that the term is antiquated. Its prevalence differs between institutional types, with institutions that have faculty unions rarely employing it but more than a third of master’s institutions without collective bar­gaining including the term in their regulations.

I included institutions that identified “moral depravity” as grounds for dismissal in this count but not those that cited only “immoral conduct,” which seemingly designates a lesser infraction. The 1940 Statement provides a definition of moral turpitude in one of the 1970 interpretive comments:

The concept of “moral turpitude” identifies the exceptional case in which the professor may be denied a year’s teaching or pay in whole or in part. The statement applies to that kind of behavior which goes beyond simply warranting discharge and is so utterly blameworthy as to make it inappropriate to require the offering of a year’s teaching or pay. The standard is not that the moral sensibilities of persons in the particular community have been affronted. The standard is behavior that would evoke condemnation by the academic community generally.

Performance-related grounds. This category includes grounds that may raise concerns from an AAUP policy perspective, in particular when such grounds are directly tied to post-tenure review or annual reviews, such as in the following example found in the sample: “Non-reappointment of a tenured faculty person may occur as a result of ‘cause,’ which shall include ‘chronic low performance,’ defined as having received two consecutive ‘Unsatisfactory’ ratings.” Overall, the use of performance-related grounds has fallen since 2000, from 23 percent to 17 percent. The central recommendation of the AAUP in this context, taken from Post-Tenure Review: An AAUP Response, is the following:

In the event that recurring evaluations reveal continuing and persistent problems with a faculty member’s performance that do not lend themselves to improvement after several efforts, and that call into question his or her ability to function in that position, then other possibilities, such as a mutually agreeable reas­signment to other duties or separation, should be explored. If these are not practicable, or if no other solution acceptable to the parties can be found, then the administration should invoke peer consideration regarding any contemplated sanctions.

The standard for dismissal or other severe sanction remains that of adequate cause, and the mere fact of successive negative reviews does not in any way diminish the obligation of the institu­tion to show such cause in a separate forum before an appropriately constituted hearing body of peers convened for that purpose. Evaluation records may be admissible but rebuttable as to accuracy. Even if they are accurate, the admin­istration is still required to bear the burden of proof and demonstrate through an adversarial proceeding not only that the negative evalua­tions rest on fact, but also that the facts rise to the level of adequate cause for dismissal or other severe sanction.

Insubordination. The analysis included related terms, such as “contumacious conduct,” under this category. The use of this term in dismissal policies has increased since 2000, from 8 percent to 12 percent. The AAUP has long opposed insubordination as a ground for dismissal, as indicated by the following passage from the report of an investigating committee:

The characterization of [the faculty members’] conduct as insubordinate would seem more appropriate to a military organization or indus­trial enterprise than to an institution of higher learning. In the academic context, allegations of irresponsibility and unwillingness to cooperate place a damper upon academic freedom.6

Rare and unusual grounds for dismissal. The follow­ing examples from the sample of grounds for dismissal are found very rarely in institutional regulations. Because they relate to the reputation, interest, or mis­sion of the institution and not to the subject faculty member’s professional fitness, all of them depart from AAUP-recommended standards:

  • “commission or omission as to any matter which reflects adversely upon the college or may jeopardize the college’s reputation”
  • “active and voluntary participation in activities deliberately and specifically designed to dis­credit the college”
  • “other improper conduct which is seriously injurious to the best interests of the university or its components”
  • “demonstrated lack of support for the mission of the university”
  • “intransigent refusal to conform to university processes or policy where such behavior places the university at risk”

Financial Exigency Policies

The AAUP explicitly recognized financial exigency as grounds for the termination of appointments in the 1940 Statement. That document, however, does not define the term. It specifies only that such a condition should be “demonstrably bona fide.” The AAUP also recognizes a bona fide program discontinuance for educational reasons, even in the absence of financial exigency, as a basis for terminating appointments.

A full set of procedural standards for both grounds, as well as definitions of the terms employed, are set forth in the Recommended Institutional Regulations. The primary concern of these standards is to discour­age an administration from using either financial exigency or program discontinuance as pretext for violating faculty members’ academic freedom.

In conducting research for this section and the next, I first identified in the handbooks and collective bargaining agreements policies for the termination of appointments based on financial and program­matic conditions. I classified as not having policies both those institutions whose handbooks or con­tracts made no mention of financial or programmatic grounds for terminating appointments and those that did but lacked any actual policies for implementing such terminations. Overall, 95 percent of institutions surveyed have financial exigency policies, and 85 percent of institutions have program discontinuance policies. Both types of policies increased in preva­lence over the past two decades: in 2000, 91 percent of institutions surveyed had policies on financial exigency, and 81 percent had policies on program discontinuance. And both types of policies are more common at bachelor’s and master’s institutions than at research institutions. With respect to prevalence, little difference exists between institutions with fac­ulty unions and those without faculty unions. In the following, the prevalence of features of these policies is calculated relative to the institutions that have such policies rather than to all institutions.

Key questions about policies for terminating faculty appointments based on financial grounds are whether they employ the term “financial exigency” to describe those grounds, whether they define “finan­cial exigency” or another term used in its place, and from what source they have taken the definition. The AAUP consistently employs the term “financial exigency” in its policy documents and provides a spe­cific definition, revised in 2013, in the Recommended Institutional Regulations. My analysis classified poli­cies that employed either the pre-2013 or the current definition as being based on the AAUP’s definition.

From the AAUP’s perspective, employing the term “financial exigency” as well as the Association’s defi­nition of that term is clearly preferable to using other terms and definitions. In the absence of a definition, the AAUP’s definition can be more easily invoked if the policies use “financial exigency” rather than another term. Thus, again, the analytical categories form a hierarchy relative to AAUP policy.

Overall, I found the use of the term “financial exigency” to be very common: 81 percent of institu­tions that have a policy that allows for the termination of appointments based on financial considerations use the actual term. There is a marked difference in prevalence in the use of the term between institutions that do and those that do not have faculty unions; the prevalence at the former is half (44 percent) of that at the latter (90 percent).

Figure 7 presents findings on the provenance of definitions of financial exigency in Trower’s study and in the present study. Compared with 2000, fewer institutions today that have a financial exi­gency policy include no definition of the conditions in which the policy can be invoked (69 percent in 2000 compared with 55 percent today). The preva­lence of the AAUP’s definition has increased from 8 percent to 13 percent, while the inclusion of other definitions has increased from 23 percent to 33 per­cent. My analysis of the provenance of definitions includes both institutions that use the term “finan­cial exigency” and those that do not, mirroring the analysis in Trower’s study.

Regulation 4c of the Recommended Institutional Regulations includes procedural standards for termi­nating appointments because of financial exigency. Trower’s study analyzed the following features of financial exigency policies; I provide for each a quoted passage from Regulation 4c to explain the AAUP’s policy, as well as comments on the analysis:

  • Notice or severance. According to the relevant provision from Regulation 4c, “In all cases of termination of appointment because of finan­cial exigency, the faculty member concerned will be given notice or severance salary not less than as prescribed in Regulation 8.” Regula­tion 8, in turn, provides for at least one year of notice for tenured faculty members whose appointments are terminated. I categorized any regulation in which notice or severance pay of some kind is required, even if it is less than what Regulation 8 calls for, as an instance of providing notice or severance salary.
  • Reinstatement. Regulation 4c provides as fol­lows: “In all cases of termination of appoint­ment because of financial exigency, the place of the faculty member concerned will not be filled by a replacement within a period of three years, unless the released faculty member has been offered reinstatement and at least thirty days in which to accept or decline it.” Any policy that provides for reinstatement, even if the time period covered is shorter (or longer) than three years, was classified as providing for reinstatement.
  • Faculty role specified. Regulation 4c calls for meaningful faculty involvement both in the dec­laration of a state of financial exigency and in the selection of individuals whose appointments are to be terminated. I categorized any policy in which the role of a faculty governance body or the faculty union was specified (even a role that departed from AAUP recommendations) as specifying the faculty role, unless that role was limited to the faculty’s merely being informed by the administration after the declaration had been made.
  • Another suitable position. The relevant provision from Regulation 4c states, “Before terminating an appointment because of financial exigency, the institution, with faculty participation, will make every effort to place the faculty member concerned in another suitable position within the institution.” Some institutional regulations called for less than “every effort,” but these institutions were still included as providing for “another suitable position.”
  • Preference for tenured faculty. The relevant provi­sion from Regulation 4c states, “The appoint­ment of a faculty member with tenure will not be terminated in favor of retaining a faculty member without tenure, except in extraordinary circumstances where a serious distortion of the academic program would otherwise result.” Only regulations that give explicit preference to the retention of tenured faculty were counted as providing preference for tenured faculty. At some institutions, particularly at those with faculty unions, the contract or handbook based such preferences strictly on seniority rather than on tenure status. These were not included as preferring tenured faculty.

Together with the overall increase in the prevalence of financial exigency policies, the prevalence of the above listed features has either increased or, in one case, stayed unchanged since 2000 (figure 8). The largest increase compared with that study was in the number of policies that specify the role of the faculty, which increased from 50 percent to 66 percent.

The prevalence of each of these features is higher at institutions that have faculty unions than at those that do not (figure 9). The differences are quite large, with the largest being 37 percentage points for reinstate­ment (63 percent versus 100 percent). In the case of preference for tenured faculty, the difference is 31 per­centage points (38 percent versus 69 percent), which does not account for the number of union contracts that give preference based strictly on seniority rather than on tenured status.

Program Discontinuance Policies

As noted in the previous section, the prevalence of program discontinuance policies has increased from 81 percent to 85 percent since 2000. The prevalence differs by institutional type, with such policies being least prevalent at research institutions, where only 75 percent of institutions have them.

The Association’s policies on termination of appointments because of program discontinu­ance specify that such a discontinuance needs to be “based essentially upon educational considerations.” Regulation 4d of the Recommended Institutional Regulations notes, “‘Educational considerations’ do not include cyclical or temporary variations in enrollment. They must reflect long-range judgments that the educational mission of the institution as a whole will be enhanced by the discontinuance.” Like Trower’s study, this study views policies that cite “academic” considerations or the outcomes of regular program review processes as involving edu­cational considerations. The AAUP does not regard policies as based on educational considerations if they treat budgetary and educational considerations equally or include only budgetary considerations. The following provision from one of the handbooks in the sample illustrates the sort of grounds for program discontinuance against which Regulation 4d aims to guard: “Financial reasons which, though they do not constitute an emergency for the college as a whole, do suggest that continuation of the program would not be in the best interests of the college.” As was the case in Trower’s analysis, when a policy did not cite any specific grounds for program dis­continuation, it was not considered to be limited to educational considerations.

The prevalence of discontinuance policies that limit themselves to educational considerations is 33 percent overall. Such policies are less common among collective bargaining institutions (26 percent) and among research institutions (19 percent) that have program discontinuance policies. The overall prevalence has declined somewhat since 2000, when it was 36 percent.

As with financial exigency policies, Trower ana­lyzed some of the procedural features of program discontinuance policies. Again, I cite for each the relevant passage from the AAUP’s Recommended Institutional Regulations:

  • Notice or severance. Regulation 4d states, “If no position is available within the institution, with or without retraining, the faculty member’s appointment then may be terminated, but only with provision for severance salary equitably adjusted to the faculty member’s length of past and potential service, an amount which may well exceed but not be less than the amount prescribed in Regulation 8.” Regulation 8, in turn, requires a minimum of twelve months of notice or severance for tenured faculty. Again, I classified any regulation in which notice or severance salary of some kind was specified, even if less than what Regulation 8 calls for, as requiring notice or severance salary.
  • Another suitable position. According to Regu­lation 4d, “Before the administration issues notice to a faculty member of its intention to terminate an appointment because of formal discontinuance of a program or department of instruction, the institution will make every effort to place the faculty member concerned in another suitable position.” Some institu­tions called for less than “every effort,” but I still included these institutions as requiring “another suitable position.”
  • Faculty role specified. Regulation 4d regards “the faculty as a whole or an appropriate committee thereof” as primarily responsible for deter­mining the educational considerations used to decide whether to discontinue a program. I categorized policies that included other specifi­cations of the role of the faculty as specifying a role for the faculty so long as the administration did not merely inform the faculty a decision has already been made.
  • Retraining. Regulation 4d adds to the provision regarding another suitable position, “If place­ment in another position would be facilitated by a reasonable period of training, financial and other support for such training will be proffered.”

Again, the prevalence of all of the analyzed features has increased since 2000 (figure 10), with policies specifying the faculty’s role in the decision-making processes increasing the most (from 53 percent to 62 percent). Providing notice or severance and making efforts toward finding another suitable position are both highly prevalent (83 percent and 76 percent, respectively). Both are more prevalent among program discontinuance policies than among financial exigency policies (where their prevalence is 74 percent and 62 percent, respectively).

The prevalence of these features differs markedly depending on the presence of a collective bargaining contract: all are more commonly found in contracts than in faculty handbooks (figure 11). The difference is largest with respect to policies that specify the role of the faculty (25 percentage points, with 82 percent of contracts specifying the role and only 57 percent of handbooks doing so).


This report has provided an overview of findings of a partial replication of a study of faculty appointment policies conducted twenty years ago. Central findings of the report are the following:

  • The 1940 Statement of Principles continues to serve as the primary source of academic freedom language in faculty handbooks and collective bargaining contracts: 73 percent of institutions with a tenure system base their aca­demic freedom policy directly on it, and more than half attribute the language to the AAUP.
  • Common grounds for dismissal for cause in faculty handbooks and contracts are consistent with the policies of the Association, and, con­versely, those that the AAUP views as problem­atic are rare.
  • Policies concerning terminations of appoint­ment because of financial exigency have become more common, occurring at 95 percent of institutions. The prevalence of the term “finan­cial exigency” in those policies differs between institutions that do and those that do not have faculty unions, with 44 percent prevalence at the former and 90 percent prevalence at the latter. The prevalence of procedural elements found in these policies has increased since 2000, with specific provisions concerning the role of the faculty increasing the most, from 50 percent to 66 percent. The prevalence of each of these procedural elements at institutions at which the faculty engage in collective bargaining is higher than at institutions without faculty unions.
  • Policies concerning terminations of appoint­ment because of program discontinuance have also become more common and can be found at 85 percent of institutions. The prevalence of program discontinuance policies that are “based essentially upon educational consid­erations” is less common among collective bargaining institutions (26 percent) than at those without faculty unions (33 percent). Again, the prevalence of all the analyzed features has increased since 2000, with the percentage of policies specifying the role of the faculty increasing the most (from 53 percent to 62 percent). All of these features are again more commonly found in collective bargaining contracts than in faculty handbooks.

To a limited extent, the prevalence of AAUP-supported procedural standards can be viewed as a proxy for how well academic freedom is protected at institutions in the population. That is, the reason that the AAUP advocates the inclusion of its policies into institutional regulations is that it believes that they serve to protect academic freedom, and thus the preva­lence of such policies provides some indication of how well academic freedom is protected. These findings do, of course, have to be tempered with the observation that administrations and governing boards have been known to disregard their own institutional policies when taking various personnel actions. Nevertheless, changes in prevalence of these policies over the course of the past two decades provide information about changes in the nature of the protection of tenure and in the climate for academic freedom. Of course, other ways to measure the climate for academic freedom should be considered and compared with the findings reported here in order to assess their usefulness.

Appendix: Methodology

The point of departure for this study was the stratified random sample of Trower’s study, which consisted of 217 institutions. The goal was to retain as many institutions as possible from the original Trower sample in order to increase the direct comparability between the two sets of results. The eight categories from the Carnegie classification system that Trower used to stratify her study’s sample—Research 1 and 2, Doctoral 1 and 2, Master’s 1 and 2, and Bachelor’s 1 and 2—are no longer reported in IPEDS. The pres­ent study instead employed the immediate successor classification system, Carnegie 2000, for which IPEDS still reports designations. The six categories in table 1 correspond to the eight previous categories and were employed here instead. For the purpose of the current analysis, each of the pairs of categories was combined into the three categories—Research, Master’s, and Bachelor’s—used throughout the report.

Of the 217 institutions in Trower’s sample, four have closed, five have merged with other institutions, and one no longer has a Carnegie classification that is represented among the categories used here. For the purpose of the current study, I replaced these ten institutions with institutions that were from the same states and shared similar characteristics. In the cases of three institutions that had merged with institutions that still belonged to the Carnegie classifications included in the sample, the merged institution was selected.

While Trower had to determine from institutional regulations the presence of a system of tenure at colleges and universities in her sample, IPEDS now collects information about the presence of a tenure system. This study used that information to restrict the sample to institutions with a tenure system. Since 2000, of the 217 institutions in Trower’s sample, four had adopted a tenure system, and three had abolished their existing tenure system. Some of the substitutions noted above also resulted in a modest change in the number of institutions with a tenure system in my sample; thus, while 196 institutions in Trower’s sam­ple had tenure, the final number in the present sample is 198. The overlap of institutions with tenure systems between Trower’s sample and this sample consists of 174 institutions (89 percent of her sample).

For this study I endeavored to collect faculty hand­books and collective bargaining agreements from the websites of all of the institutions and faculty unions in the sample. Twenty-nine institutions, however, restricted access to their handbook or contract by making it available on-campus only, often through a proprietary human resources portal. I contacted faculty members and administrators at these twenty-nine institutions with a request that they provide a copy of the current handbook or contract; fourteen agreed to do so. I substituted similar institutions for the fifteen nonrespondents, using information obtained from IPEDS to identify comparable insti­tutions. Although substitution is not generally the preferred mechanism for addressing unit nonresponse in sample surveys, the institutions that declined to provide their regulations for this study differed by institutional type from the rest of the sample, thereby reducing the size of some of the strata to such an extent that, without substitution, would have affected variance estimation and thus margin of error. Of the twenty-nine institutions that restricted access to their regulations, twenty-three were bachelor’s institutions and six were master’s institutions. Five of the master’s institutions submitted their regulations, and thus fourteen of the fifteen nonresponding insti­tutions were bachelor’s institutions. Substituted units were compared with those units that had submitted their restricted institutional regulations in order to determine whether the two groups differed systemati­cally, which I found not to be the case.

I analyzed the regulations in the sample using qualitative analysis software, and I analyzed the results with a statistical software package. Although Trower’s study used a stratified sample in which she selected institutions from each stratum with unequal probabilities (in other words, the sample was not self-weighting), results reported in her study were not weighted. In order to improve the accuracy of the esti­mates, I weighted results from her study reported here whenever it was possible to do so, and I weighted the results of the present study with design weights and with post-stratification weights based on the preva­lence of institutional control in the population.

Estimates of prevalence in the population made on the basis of a sample have a margin of sampling error. The margin of error depends on the size of the sample and of the prevalence itself. For a sample size of 198 (the overall sample of institutions with a tenure sys­tem) it is +/- 6.35 points when the proportion reported is 50 percent, which is when the margin of error is largest for a given sample size. Thus, for example, the estimate that 52 percent of institutions in the popula­tion have the 1940 Statement with attribution in their regulations has a 95 percent confidence interval of 45.5 percent to 58.4 percent. The margin of error is larger when statistics are reported for subpopulations (such as by Carnegie classification, collective bargain­ing status, and so forth).

Senior Program Officer and Researcher, AAUP

1. See Hans-Joerg Tiede, “Introduction: AAUP Policies and Their Effective Use,” in AAUP, Policy Documents and Reports, 11th ed., (Baltimore: Johns Hopkins University Press, 2015), xiii–xxi, for an overview of AAUP policy for the purpose of incorporating it into institutional regulations.

2. Cathy Trower, ed., Policies on Faculty Appointments: Standard Practices and Unusual Arrangements (Boston: Anker, 2000).

3. For additional information, see AAUP, “Faculty Handbooks as Enforceable Contracts: A State Guide,”

4. IPEDS does not collect information about whether faculty groups at an institution engage in collective bargaining, and I am not aware of any other current estimates of this prevalence. Thus, the numbers reported here, as well as all of the findings concerning faculty personnel policies in this report, estimate prevalence in the population on the basis of the sample. Such estimates have known margins of sampling error, which are briefly discussed in the appendix on methodology.

5. Commission on Academic Tenure in Higher Education, Faculty Tenure: A Report and Recommendations (San Francisco: Jossey-Bass, 1973).

6. “Academic Freedom and Tenure: Illinois College of Optometry,” Academe, November–December 1982, 17a–23a.


Report Category: