Protecting Academic Freedom or Managing Reputation? An Evaluation of University Social Media Policies

Abstract

University social media policies appear to favor institutional reputation over the tenets of academic freedom upon which research and teaching are based. A content analysis of social media policies of 82 doctoral-granting research universities found that policies used language and concepts that restrain online faculty speech. Two-thirds of policies are overseen by marketing departments rather than by committees that include faculty members as recommended by the American Association of University Professors. Contradictory boundary logic presented double binds for faculty who were enjoined to be brand ambassadors, creating tension for them and creating paradoxes for institutions whose social media policies contradict their commitment to freedom of research and teaching.

The development of social media, which allow university faculty to publish professional and personal opinions that are immediately available to specific followers or the entire social media audience, exposed a gap in university policies governing personal and professional extramural speech. Facebook and Twitter were launched and available to anyone with a free account in 2006. Six years later, universities began taking public action against faculty whose social media posts had the potential to impact institutional reputation. In 2013, Chicago State University sent two professors a cease and desist order in an attempt to shut down a blog that was critical of the school's administration.¹ In a highly publicized case, David Guth, a professor of journalism at the University of Kansas, was suspended following an emotional tweet he posted in response to a mass shooting at the Washington Navy Yard.² Though the incidents happened within months of each other, the resolutions were very different. Guth's tweet led to a new social media policy (SMP) for Kansas universities that gave universities the right to suspend or fire university faculty or staff for social media posts deemed “contrary to the best interests of the university.”³ The Chicago State case was settled in 2019 in favor of the two professors who ran the blog. In addition to a cash settlement, the university agreed to rewrite policies governing computer usage.⁴ Nor are these cases isolated. Steven Salaita saw his offer of a position at the University of Illinois–Urbana-Champaign withdrawn after tweets that drew condemnation for their vulgarity, epithets, and naked anti-Semitism.⁵ In 2019, the provost at the University of Indiana, Bloomington, called a professor's views “loathsome, stupid and ignorant,” but added that his tweets were protected speech.⁶ These divergent responses indicate fundamentally contrasting philosophies about whether faculty comments on social media represent their own opinions or are a reflection of the university. And if a university decides that a professor's comments are indeed reflective of the institution, are the posts still protected under the banner of academic freedom, or does the university's need to protect its reputation take precedence?

Academic freedom is a broad principle that protects the right of faculty to research, publish, and teach material that may be unpopular, controversial, or even wrong. While the policy is widely accepted, at least in principle, universities must strike a balance between faculty's rights to academic freedom, schools' needs to protect their reputations, and the opinions and demands of internal and external stakeholders that interject themselves into exchanges on social media. Communication policies are one of a university's strategies for managing competing interests. SMPs extend this mediation into the digital age. In their study of boundary regulation among major corporations, Banghart, Etter, and Stohl⁷ found that corporations increasingly use invasive boundary logic in SMPs that constrains employee speech, self-expression, and relational engagement, and that places the financial and public relations interests of the organizations ahead of fostering an open environment. As universities continue to restructure their governance models toward a business orientation, reputation management becomes more important as controversies can deter major donors and attract unwanted media and political attention. However, universities are not corporations, where management is answerable only to stockholders and a board of directors. The goal of a university is to foster research and teaching in an open environment. SMPs that limit that function may compromise their missions to educate and create knowledge.

We investigated the conflict between academic freedom and reputation management that plays out in SMPs. Using qualitative content analysis outlined by Zhang and Wildemuth,⁸ we found that faculty, whose research and speech are protected by academic freedom, are also constrained by SMPs that restrict their ability to speak freely and require that their social media be consistent with university messages. We are aware that faculty are no different from anyone else in that their speech can be controversial and even hurtful, and that expressing these views on social media can have foreseen or unforeseen consequences within the university walls and outside of academia. Nevertheless, the ability to pursue research agendas that may be unpopular is central to the mission of research universities and necessary to developing knowledge that advances our understanding of the world, influences public policy, and enhances our lives. We begin with an overview of research related to academic freedom, reputation management, tensions, and university SMPs. Next, we discuss the Internet search that led to our corpus, 82 SMPs of R1 research universities, and the methods we used for content analysis. Then we describe tensions between SMPs and academic freedom and conclude with implications and recommendations for further research in this area.

Literature Review

Academic Freedom

The ideals of academic freedom were articulated more than a century ago by the American Association of University Professors (AAUP).⁹ This document considered three activities that need to be protected: research, teaching, and speech outside the laboratory or classroom. The organization has not wavered in this approach, but it has updated it twice to reflect the changing media landscape. A 2013 policy statement affirmed, “Academic freedom, free inquiry, and freedom of expression within the academic community may be limited to no greater extent in electronic format than they are in print.”¹⁰ By 2013, several incidents of faculty being censured or suspended for social media use had made headlines in local and national media. In each case, the AAUP defended the speech not for its content but for the right of faculty to freely express themselves even in areas of political speech or opinion. The 2013 policy statement specifically addressed faculty use of social media for extramural speech and recommended that institutions include faculty members when developing SMPs. Policies must “recognize that social media can be used to make extramural utterances and thus their use is subject to Association-supported principles of academic freedom, which encompass extramural utterances.”¹¹ At the same time, the policy recognized universities' responsibility to create SMPs, although the AAUP's concerns included acceptable use, copyright protection, and access to electronic communication technologies. The document noted that although SMPs were not yet the norm for colleges and universities, “at institutions where such policies exist, the focus is frequently on the university's reputation and not on the faculty's academic freedom.”¹²

While academic freedom and the US Constitution's First Amendment protections overlap, there are significant differences that affect faculty.¹³ The First Amendment protects everyone from government intrusion on public oral, written, and now, virtual speech. In 1956, the American Civil Liberties Union (ACLU) stated that university faculty do not give up their First Amendment rights when they speak as private citizens.¹⁴ Academic freedom protects faculty speech more widely within and outside the university setting, covering their research, teaching, and speech inside and outside the classroom or research setting. However, the principle does not have legal force, giving universities greater freedom to deal with perceived affronts to reputation within its walls. Even so, violations of the First Amendment can open the university to lawsuits by aggrieved parties.

Risk, Reputation, and Brand

Risk can be described as the potential for loss due to changes in financial, operational, legal, or regulatory conditions, which can damage an organization's reputation.¹⁵ As a result, organizations spend considerable human and financial resources to manage risk, extending the reach of internal controls into every aspect of the organization.¹⁶ For universities, the desire to minimize risks to their reputation can lead to friction between administrations and faculties (and other stakeholders), and controlling risk may come at the expense of academic freedom.¹⁷

Reputation has been described as people's perception of an organization's competence and its standing in the eyes of stakeholders.¹⁸ This definition underscores the complexity for higher education institutions whose internal and external stakeholder groups have diverse interests that may compete with one another. Universities have a legitimate interest in maintaining their good names. The quality of their reputations affects their ability to compete in the global arena for students, faculty, research dollars, and donors.¹⁹ Indeed, university reputations are on display in annual rankings in publications like U.S. News and World Report in the United States and QS World University Rankings internationally. As the ability of faculty to reach wider audiences expands through social media, universities have become more invested in monitoring and controlling the online messages of employees and others.

Researchers have linked a university's reputation to identity, brand, and image, described as reflections of stakeholder opinion that create greater awareness and recognition.²⁰ Creating, enhancing, and maintaining these university assets are functions of public relations and corporate communications departments; safeguarding the university's reputation is often the purview of these departments as well.²¹ The danger is that when reputation is viewed as a public relations issue, the response can become a matter of issuing press releases, mediating access between faculty and external media,²² and governing image and brand assets like logos, brand colors, and mascots,²³ instead of grappling with issues of content, ethics, and rights.

The broad reach of social media has been both a boon and a challenge for branding. There is a positive correlation between successful brand outreach and management on social media and competition for students.²⁴ The popularity of blogs, Twitter, and other platforms can lead to tensions between academics, who use them to communicate directly with various audiences, and university public relations professionals, who seek to control the information flow as a means of supporting the university's brand. In the words of one academic whose controversial research findings caused a social media firestorm, “the Internet and social media can be a powerful instrument in the hands of those who want to control academia.”²⁵

University SMPs

Social media are Internet-based applications that allow the creation and exchange of user-generated content.²⁶ They are a well-used method of communication within universities; by 2011, it was estimated that 90% of teaching faculty in higher education used social media.²⁷ SMPs are a governance framework for balancing user participation in online media with control of corporate interests.²⁸ In 2013, Vaast and Kaganer reviewed SMPs of 74 corporations, nonprofits, and universities. They coded the policies based on common characteristics that included defining the parameters of social media, guidelines for dealing with blurred boundaries between professional and personal lives, advice on what to post or not post, suggestions for creating online communities, editorial style, issues surrounding personal and organizational identification, warnings about the persistence of social media, advice on how to represent oneself, mentions of governance and support, enforcement of guidelines and links to established policies including information and communication technology (ICT) policies, penalties for infractions, and paths to content and governance authorities.²⁹ These codes formed the basis of the coding schema used in our study.

Despite SMPs' effect on various stakeholders, there are few studies on university SMPs. In 2012, McNeill investigated the SMPs of 14 universities in the United Kingdom whose policies were available online. He found that half of these policies originated within departments dedicated to marketing and noted the primacy of reputation management, which he found in two-thirds of policies. References to reputation, integrity, or the interests of the university were found in preambles or mentioned in other prominent places. Compliance was encouraged or assured via policy levers (mechanisms that govern how or when users can access or use social media) and disciplinary actions in university policies that controlled the behavior of social media users. McNeill noted that marketing terms like “promoting,” “brand,” and even “brand management” were standard in the policies. Language often sounded more collegial than the legalese of formal policies, even when the levers could trigger punitive responses.³⁰ This emphasis on language that sounds friendly but is intended to control behavior has been seen in corporate policies as well.³¹

In their comprehensive research study of 822 college and university policies available online, Pomerantz, Hank, and Sugimoto noted the prevalence of language that elevated protecting reputation over the ideal of academic freedom. This study used similar (but not identical) categories to Vaast and Kaganer, grouped into three categories: appropriateness of posts (e.g., what to post and what not to post), representing the institution (e.g., branding and identifying oneself as a member of the institution), and compliance with the law (e.g., hyperlinks to other policies within the organization and references to conduct toward legally protected categories of students, coworkers, and people outside the universities as mandated by state and federal laws).³² Neither McNeill nor Pomerantz et al. measured academic freedom or First Amendment protections in their studies, but Pomerantz et al. concluded, “Future policies will need to navigate the balance between academic freedom and institutional branding with care.”³³ McNeill stated, “Policies constrain as much as they enable our possibilities for action.”³⁴

Double Binds, Paradoxes, and Tensions

As work grows in complexity, needs and interests in the workplace often conflict. Scholars have noted the existence of double binds, paradoxes, and tensions with which workers at all levels must contend.³⁵ Double binds are characterized as the dilemma felt by employees faced with contradictory messages and expectations.³⁶ Employees are stuck in a no-win situation as they attempt to navigate between competing rules.

Paradoxes are contradictory principles or rules that stand in direct opposition to each other. Obeying one makes obeying the second impossible.³⁷ For example, a directive for social media users to “be yourself” while representing the university presents a paradox for users whose opinions do not represent the views of the organization. Paradoxes may arise when employers intend to empower workers, but the policies they design have the opposite result, creating distrust among the employees.

Tensions are clashes of ideas, principles, and policies that embody contradictory logic or competing demands.³⁸ They are the byproducts of processes that arise when organizations or their employees are pulled or purposely moved in different directions and now must achieve two opposing goals. Any choice along the continuum results in tension within the organization, which stakeholders must work to resolve. When policies lean far enough to one end of the continuum, it becomes more difficult for the organization to accommodate those on the other side.

Corporate policies are often used both to define our work lives and to permeate the physical, temporal, and relational boundaries that separate them from our personal ones. Banghart et al. and Boswell and Olson-Buchanan describe how boundary logics and control in corporate SMPs serve to inhibit employee speech and self-expression, often intentionally.³⁹ SMPs were shown to use strategic ambiguity to construct invasive and contradictory logics that privileged employer interests over those of employees.⁴⁰ Policies that state “you are always an employee” exemplify invasive boundary logic that appropriates employees' identities outside the workplace and brings them under the jurisdiction of the organization. The ambiguous language has both positive and negative effects for corporations. While creating permeable boundaries that allow corporations to regulate speech even when employees are not in the workplace, SMPs also generate confusion and, often, conflict and stress.⁴¹

Weber demonstrated that corporate SMPs contained contradictory policy logic that created a paradox for employees. His examination found that SMPs “establish potentially contradictory expectations” and “ambiguous relationships” between writers and their corporations.⁴² We could find no similar studies that analyzed university policies for paradoxes and tensions, although we found references to these contradictions in our reading of current events and background materials. The Chronicle of Higher Education and the AAUP's blog have published work about the subject since the Guth event. The Chronicle called the Kansas SMP “an affront to First Amendment freedom of speech and academic freedom.”⁴³ The AAUP's blog explained that academic freedom is always outweighed by other factors in the new policy, including the prohibition of blogging during “working hours,” using a university computer, or mentioning anywhere on your blog that you are a professor at a Kansas University.⁴⁴ Legal journals have also noted this tension. Solberg argued that the language of SMPs might deter faculty from acting as individuals and can be a constraint on faculties' freedom of expression.⁴⁵

We investigated SMPs and how they bring institutional image management into conflict with academic freedom.⁴⁶ We contend that faculty, whose research, teaching, and internal and external utterances are covered by universities' commitment to the principles of academic freedom, are limited at the same time by policies that restrain their ability to post and require their adherence to institutional brand messaging. In this context, we address the following research questions:

• RQ 1: How do universities encourage and constrain the use of social media through policies?

• RQ 2: What tensions, paradoxes, and double binds are revealed in these policies?

Methods

This study utilizes qualitative content analysis to investigate the conflict between academic freedom and reputation management within university SMPs. We examined SMPs from an initial sample of 116 public and private universities and colleges designated “R1 Doctoral Universities” under the Carnegie Classification of Institutions of Higher Education^® (CCIHE). These are institutions that awarded at least 20 research/scholarship doctoral degrees (exclusive of professional practice degrees like JDs or MDs) during the year they were reviewed by the CCIHE and were rated as having the “highest research activity.” “Special focus institutions” and “tribal colleges” were not included in the classification either.⁴⁷ We believed that the nature of the R1 universities and colleges (prominent, research-focused, and well-established) meant they valued academic freedom and likely encouraged, or at least tolerated, research that might veer into controversy. They understood the potential benefits and pitfalls of social media, and had the resources available to craft policies and monitor compliance. Additionally, there was a reasonable expectation that their policies would be available online, given that many are public schools and must meet the needs of stakeholders with vastly different needs and interests.

The data were collected as part of a larger project examining policies within the context of higher education organizations. Eight graduate students who were members of a seminar focused on work and communication technology searched for the policies. They used Boolean search parameters on Google and the universities' home page search engines to locate documents. Search terms used were “social media policy,” “social media guidelines,” and “social networking.” Students were asked to include both policies and any social media guidance, including sections of student or employee handbooks, overall policies for acceptable use of technology or the Internet, and public relations, marketing, or communications guides in their reports. Each webpage or downloaded document was saved as a Portable Document Format (PDF) or MS Word document. The webpage address was inserted into the first page of each document, and each file was named according to a convention established in the instructions to the students. Each student also submitted a tracking document in MS Excel that included institution names, document names, and the corresponding web addresses for each document.

Each of the 116 R1 universities was assigned to two students, and they were instructed to spend at least 20 minutes searching each school's website. If neither student was able to locate a relevant policy at that time, we concluded that no such policy was publicly available. (There were some instances where students found handbooks or other content stored on a university intranet that was available only with a university log-in ID and password. We were unable to examine these documents.) In total, 427 documents were returned (including duplicates found by both researchers assigned to a single university) yielding 358 unique documents. We first identified policies created for the entire institution (versus secondary policies published by a single school or unit within the university). These institutional-wide policies were all kept in the sample, regardless of whether they were located by one or both students assigned to that institution. Then we examined secondary policies. Policies located by both students were considered to be significant due to their greater prominence; they remained in the sample. Some institutions had multiple policies that remained in the sample. We consolidated these into one document so that the level of analysis was the university rather than specific policies.

Though we found social media–related policies or guidelines for all 116 R1 universities, not every university had an institution-wide SMP or even a school- or unit-specific policy that was found by both researchers assigned to that university. As a result, our final sample consisted of consolidated policy documents from 82 universities and colleges. Because we focused on publicly available university policies, the names of the universities were retained in our analysis and discussion for clarity and context; names of individuals were removed. (See  Appendix A for a complete list of universities included in our sample.)

Analysis

For our analysis, we used NVivo 12 software, which allows coders to select, store, categorize, and analyze text from electronic documents. Both authors reviewed 20 policies (approximately 25% of the sample) using the previously mentioned codes defined by Vaast and Kaganer. The purpose of this was threefold. First, we were training to ensure that our interpretations were consistent with Vaast and Kaganer's. Second, we determined that even though the core ideas of several of Vaast and Kaganer's codes matched our points of interest, they required different titles in order to better align with the naming conventions used in our sample (e.g., their “editorial recommendations” became our “best practices”). All changes are noted in Table 1. Finally, we also used this process to develop additional codes specific to higher education and our research questions that could be used to supplement Vaast and Kaganer's more general, corporate-oriented codes. These included references to academic freedom, freedom of speech or First Amendment rights, reputation management, and the role of social media users within organizations. We addressed the areas where we thought tensions might exist. For example, the code “roles” highlighted language such as “In large part, these Guidelines (sic) are addressed to employees who use social media as part of their jobs to promote their schools, programs, and departments.”

Once this augmented set of codes was created, we independently examined each policy document to identify examples of the characteristics described by the codes. We used a present/not present code to note whether the feature was found in the policy. We met throughout the coding process to compare examples, discuss questions, and resolve discrepancies. Some characteristics, such as “roles,” had subcodes (e.g., professional, personal, professional and personal). We also found that some policies had more than one subcode present within a given attribute. When we compared findings, we coded a particular characteristic as present if we both agreed it appeared at least once within the policy. These examples were the basis of our answer to RQ1: they displayed how university policies could encourage and constrain social media use.

Once we agreed on examples found in the texts, we re-sorted the passages of texts selected in NVivo to create individual documents by code and performed independent content analyses to ascertain whether tensions were present in the documents. Content analysis gave us a richer description of the data and provided information on the role that documents and other media play in our understanding of communicative processes and human behavior.⁴⁸ We looked for language similar to that of the policy levers described by McNeill⁴⁹ and defined by Gibbs et al.⁵⁰ When language indicating tensions or double binds was found, it was noted by the coders. We then reviewed the annotated NVivo documents to affirm our agreement. A complete description of codes, including examples, can be found in  Appendix B.

Results

Encouraging Use through Organizational Support, Value Creation, and Fostering Community

Of the twenty-four characteristics we coded, only two could be explicitly categorized as encouraging. Organizational support of social media (M = .81, SD = .4), which was found in 81 percent of the policies reviewed, generally took the form of references to resources for social media managers, including contact information for specific university employees or units. For example, Boston University's policy recommended that social media users “Coordinate and get off to the right start. Take advantage of the University's resources and expertise concerning social media. You should coordinate internally with Public Relations to learn the ropes before starting to use social media.” These policies acknowledged the potential for value creation through social media (though these were very frequently immediately followed by a warning about the pitfalls of its use). For example, one policy stated, “Social platforms provide some of the quickest and most direct ways to share information, but with their power to inform also comes institutional risk.” Language that encouraged users to generate and sustain a sense of community was present in the policies of 55 percent of institutions (M = .55, SD = .50). For example, the University of Southern California exhorts users to “Engage in a two-way [sic] conversation with your followers and strive to create a positive community.” The University of Georgia avows that “Social media provides a place to foster community and conversation. Adding value is good when on topic and in moderation.”

Constraining Policies: What Not to Post, Identify Oneself, Privacy Concerns, Follow Established Rules

Far more prevalent were attributes that placed limits upon the user. All of the most common codes appeared in three-quarters or more of the policies examined. They included what not to post (M = .81, SD = .4), as in Cornell University's policy, which states, “The university does not permit social media messages that sell products or promote commercial, political, or other ventures.” Posts that might run afoul of other university policies are forbidden; the Georgia Institute of Technology warned, “Do not post material that is harassing, obscene, defamatory, libelous, threatening, or embarrassing to any person or entity. Do not post jokes or comments based on an individual's gender, sexual orientation, race, ethnicity, age, or religion.” Social media users are instructed to identify themselves (M = .73; SD = .45) and state that their views are personal and not representative of the university. Tulane University recommended, “On personal sites, identify your views as your own: If you identify yourself as a Tulane faculty or staff member online, you must make it clear that the views expressed are not necessarily those of the institution,” whereas University of California, San Diego, said, “Identify yourself. Be transparent. No matter when or where you jump into the conversation online, speak in your own voice and clearly identify yourself and your affiliation with UC San Diego. Remember you are your area's spokesperson.”

Other policies dealt with protecting confidential or private information (M = .76, SD = .42), including personnel data, information protected by the Family Educational Rights and Privacy Act (FERPA), Health Insurance Portability and Accountability Act (HIPAA), or other legally protected or confidential information: copyright, trademarks, and the proprietary information of third parties or universities themselves. Also frequently included were directives to follow established rules (M = .72, SD = .45), for example,

Keep in mind other Clemson University policies: This document does not affect other university policies that might apply to the use of social media, including, but not limited to the Harassment policy, Computer Use policy, Workplace Violence policy, and University Web Policy. If your use of social media would violate any of the Institution's policies in another forum, it will also violate them in an online forum.

Document Descriptors: Titles, Roles, and Policy Origination

Our coding process also identified the office of policy origination, the roles of the social media user covered by the policies, and the title of the document. At some point, all of the institutions reviewed referred to a portion or the entirety of their document as “guidelines,” although only 71 percent used that term in the title (M = .71; SD = .45) Slightly less than one-third of the schools reviewed used the term “policy” in addition to “guidelines (M = .31; SD = .46);” a small percentage opted to use language such as “resources,” “criteria,” or simply “best practices” (M = .07; SD = .26). This use of terminology aligns with McNeill's and Weber's evaluations of friendlier language being used to mask constrictive policies. In terms of user role, 64 percent of SMPs addressed social media use via professional or university-affiliated accounts, for example, those run by an employee (or sometimes, a student) for a unit, department, or organization at the university (M = .64; SD = 48). Approximately 40 percent of policies covered both professional and personal accounts (M = .38; SD = .49); only 6 percent covered only personal social media accounts (M = .06; SD = .24). In some cases, the policy explicitly listed the office responsible for its creation. In others, we assumed the office whose webpage was hosting the policy was the policy author because this followed the pattern established by the explicitly named offices. The assumed authors also tended to guide social media producers to that same office if they had questions. The offices from which the policy documents originated were three times as likely to be the marketing, public relations, or communications departments as they were to be university administration or “other,” for example, external relations, human resources, or student affairs. With 55 percent of the policies originating in a marketing or similar office, this exceeds the percentage found in McNeill's 2012⁵¹ study (M = .55; SD = .50).

Tensions between Reputation Management and Academic Freedom

Also exceeding McNeill's results were references to reputation management—in our sample, 73 percent of policies contained language concerning the institution's reputation or image (M = .73; SD = .45). An example found in the University of Texas at Arlington's policy is “Know your role as representative of UTA. Keep in mind your association with UTA at all times when engaging in social media conversations. Ensure that your content is accurate, positive, and consistent with how you wish to present yourself and the institution to peers, colleagues, students, alumni, and other constituents.”

In contrast, two salient codes (both to this study and presumably to the faculty of the universities and colleges), those of academic freedom (M = .32; SD = .47) and freedom of speech/freedom of expression/First Amendment rights (M = .24; SD = .44) were present in fewer than one-third of the policies. Those schools whose policies included them frequently linked these qualities to values consistent with ideals of civility. For example, George Washington University states, “It is the policy of the university that the use of social media be consistent with its commitment to academic freedom and university values and promote thoughtful discourse on appropriate matters.” Similarly, the University of California, Riverside, avers that “Freedom of expression, the right to agree and disagree, the spirit of inquiry, and free exchange of thought are essential in an open society. As part of society, each institution of higher education must help maintain these rights for all who are a part of its concern.”

Policies that connect social media postings and reputation management generate a double bind for faculty whose opinions could be seen by stakeholders and other interested parties as undermining reputation. The conflicting policies exemplify a paradox of university SMPs: can a university uphold the canons of academic freedom and still require professors to consider the university's reputation? One school included language that acknowledged its commitment to academic freedom (“Duke University maintains its strong commitment to academic freedom in these channels”), then later stated, “When you post to social networks from an official Duke University account, you are representing the Duke brand.”

Even a disclaimer on a personal social media account is not sufficient to break ties with the university. “If readers see your affiliation with Duke, they may read your words as representing the university, even if you provide a disclaimer saying your views are your own,” declares Duke University. Cornell University's SMP clearly states its priorities. “The university recognizes the value of social media platforms for a range of business goals and must balance its support of social media with the preservation of Cornell's brand identity, integrity, and reputation.” University employees who are authorized to use social media at work are warned to “maintain Cornell's brand identity, integrity, and reputation while minimizing potential legal risks.” Boston University's policy also prioritizes personal reputation to the university. “On social media, employees should be guided by an even heightened concern for protecting their own reputations, the reputation of the University, its schools and departments, and the reputations of others, such as co-workers and students.” Once again, adding a disclaimer does not separate a faculty member from the school. “Even with such a disclaimer, remember that inappropriate postings may reflect poorly on the University and you as an employee.”

Oregon State University tells its faculty that “Though Oregon State is committed to having a strong social media presence institution-wide, it's important to ensure that messages from faculty, staff, and students are strategic and consistent with the university's overall mission and brand.” Faculty at Stanford University are reminded, “All of your posts, comments, and actions on social networks have the ability to affect the reputation of the university as well as other individuals affiliated with Stanford.” Whether reminders are collegial (“Remember that you are representing your organization and the Institute”) or more direct pronouncements like “Inappropriate postings may reflect poorly on the University,” linking social media posts to reputation is a lose–lose situation for a faculty member whose research may be deemed controversial. An SMP that requires faculty to ask themselves “Does this have potential to become a public relations issue?” creates a paradox, and subsequently mistrust, for faculty who understand that their research may generate adverse publicity, but believe that publishing advances knowledge in their field or who feel uncomfortable serving as “brand ambassadors” for their schools.

Boundary Permeability as a Double Bind

Boundary permeability was seen in many policies, creating another double bind that ensnares faculty members. Policies enjoined users to be transparent when posting, but also reminded them that they represented the institutions even in their private lives. Faculty can no longer turn off their work selves and post freely as individuals; they are reminded that “Everything you post is a direct reflection on the University of Florida.” Disclaimers, commonly seen on blogs, Twitter feeds, and other social media posts, are not sufficient for many schools. Even when employees were posting with disclaimers, universities reminded them that they represent the institution. “Faculty, staff, and students should always be aware of how they identify and present themselves to the public via their own accounts as they may be seen as de facto Rutgers authorities.” Some universities also assumed that when roles conflicted, users would continue to represent the organization. “In order to leverage the full strength of the university's name, fame, and reputation, we recommend that your social media presence be consistent and complementary to our overall brand” (UC Santa Cruz). The University of Iowa warned about damage to “the individual's and/or the university's reputation,” stating “The Office of Strategic Communication trusts—and expects—team members to exercise personal responsibility in the use of social communication both for official tasks at work and while away from work.” In these cases, faculty members can never turn off their professional roles and become private citizens. In fact, the more renowned a professor, the more intense the double bind and the resulting tension as they attempt to reconcile professional and personal lives and the tenets of academic freedom with the imperative to conform to a university brand.

In summary, we noted that the tension created for faculty that insists they prioritize their employer's reputation over their roles as researchers and teachers is both a personal double bind and a paradox for universities whose policies and practices may conflict with one another.

Discussion

A content analysis of the SMPs of 82 R1 universities showed that reputation management is a significant focus of university SMPs and that they contained language that constrained employee speech. Macnamara and Zerfass described SMPs as an attempt by management to maintain control.⁵² The attempt to regulate speech through SMPs generates tensions between two foundational credos of universities: they are places where research ideas flourish and ideas are encouraged, while at the same time, they are institutions with legions of stakeholders to whom they must answer and with reputations to protect. While university policies that favor reputation management may not raise undue concerns among students or administrative staff, independence from outside influences is considered necessary for academic research.

Role permeability puts faculty into a double bind as they grapple with policies that expect them to be an employee even when they are at home. Banghart et al. noted that corporations might strategically use ambiguous boundary logics as a way to shape SMPs.⁵³ They suggested that the resulting stress, burnout, and lower job performance of double binds may be compounded when boundary permeability is the result of corporate policies. Our study did not investigate those issues, but we confirmed the intentionality of this practice. We also affirmed the findings of previous studies that universities view policies as a marketing tool. Approximately two-thirds of the documents originated in the marketing department and were overseen by marketing or public relations staff functions.

In addition to the language within the policies, the fact that online speech is now subject to different rules forms a final tension. Faculty are not asked to become de facto brand ambassadors in any other aspect of their roles. The doctrine of academic freedom gives them tremendous freedom in the classroom even when their subject matter or teaching methods are controversial. They can publish in peer-reviewed journals and popular print media or appear on broadcast media with limited concern. It is only online that their speech is subject to oversight by others.

Conclusions, Scholarly Implications, Limitations, and Future Study

To our knowledge, this study is the first to categorize the paradoxes and tensions of university SMPs for faculty. It sustains Weber's findings of constrained agency and employee tensions and links them to the university setting. We also affirm both the AAUP's statement that risk and reputation management guide the development of university SMPs and the observation of Pomerantz et al. that the language of university SMPs privileges reputation and brand management over academic freedom. These findings highlight an area of potential discomfort for those who develop these policies and those who must live with them. Academic freedom is intended to create a space where ideas can be freely explored, independent of political or social whims. It is the bedrock of academic research, which in turn, is the mission of a research university. However, the reality of university life is that research and communication are never free from conflict. Researchers' academic interests are political choices as much as scholarly and personal ones. Their opinions, whether expressed in journals or on Twitter, can evoke strong opinions. University stakeholders and other interested parties may regard scholars' utterances as reflective of the university. As universities compete not only for students but for research dollars, rankings, and prestige, it is not surprising that leadership stakeholders want to control communication to uphold the school's image. The competition leads to tensions for faculty, but it also creates paradoxes for universities as their decision to privilege reputation management over mission may eventually rebound to their detriment.

After the Kansas Board of Regents' decision to curtail academic freedom was announced, some professors at the University of Kansas worried whether they would be at a disadvantage in attracting research faculty.⁵⁴ Over time, the university's mission could be compromised by the unwillingness or inability of faculty to pursue research. In affirming the tension between reputation management and academic freedom and the trend to favor the former over the latter, we ask not whether but how university governance affects the mission and work of the institution. We believe this is an essential area for communication scholars to investigate, as these processes are communication-driven and have implications for several areas of our discipline.

Our findings of invasive and logically contradictory regulatory activity by universities further the work of Banghart et al. by demonstrating how these contradictory rules are used in university SMPs to create permeable boundaries that blur work and home spheres for research faculty. We are not aware of studies that investigate the communicative processes used by researchers to resolve the double binds and tensions in academia. It is possible that the tension for academics may be complicated by the reality that their reputations are closely connected to those of their universities. Whether this is true and how that affects their recognition and resolution of tension is an area worthy of further research.

This study was limited by our ability to find policies publicly available on the Internet. We were unable to see behind firewalls, which might have revealed more recent, more detailed, or more nuanced documents. Additionally, this study looked at only one section of the Carnegie Classification. The 116 schools in our pool represent only a fraction of all the degree-granting institutions in the United States. Additionally, R1 universities have a special interest in research, which may make them more sensitive to reputation management that could affect present or future research partnerships. Future research can and should investigate whether these tensions occur in other categories of the Carnegie Classification.

In confining our study to questions of academic freedom and reputation management, we did not investigate the authorship of policies. We became aware that some policies were created by committees, as suggested by the AAUP in its 2013 statement, but it was not a focus of this study. The AAUP statement suggests that the organization believes that the input of faculty will affect the final policy. Future research should be undertaken to determine whether there is a relationship between authorship and the tensions outlined in this study. Finally, as our methods section noted, the two authors reviewed coding several times and refined the codebook as necessary. We did not note the Cohen's kappa during the process, which, in retrospect, is a limitation of our methodology.

Policies are a reflection of organizational priorities. The tension and paradox identified in our article reflect the conflicting values of academic institutions as they navigate between their roles as leaders in the creation of knowledge that furthers our civil society and organizations that must be responsive to disparate stakeholder groups. How they are resolved will have implications for how universities see themselves and how we as professionals and private citizens view them in return.

Acknowledgments

The authors are grateful to Dr. Craig Scott and two anonymous reviewers for their guidance and comments, and to the following Rutgers University students for their assistance collecting policy data for this research: Philip Claghorn, Angelica Flynn, Elizabeth Matusiak, Janet McGurty, Jenna Moschella, and Courtney Stevenson.

Footnotes

Bibliography

APPENDIX A: UNIVERSITIES INCLUDED IN THE ANALYSIS OF SOCIAL MEDIA POLICIES

• Arizona State University

• Boston College

• Boston University

• Brandeis University

• Brown University

• California Institute of Technology

• Carnegie Mellon University

• Case Western Reserve University

• City University of New York (CUNY) Graduate School and University Center

• Clemson University

• Colorado State University

• Columbia University

• Cornell University

• Duke University

• Emory University

• Florida International University

• Florida State University

• George Mason University

• George Washington University

• Georgia Institute of Technology

• Harvard University

• Indiana University Bloomington

• Iowa State University

• Kansas State University

• Louisiana State University

• Massachusetts Institute of Technology

• Michigan State University

• New York University

• North Carolina State University

• Northeastern University

• The Ohio State University

• Oregon State University

• Rice University

• Rutgers University–New Brunswick

• Stanford University

• Southern Illinois University

• Stony Brook University, SUNY

• Temple University

• Tulane University

• University at Albany, SUNY

• University at Buffalo, SUNY

• University of California, Irvine

• University of California, Los Angeles

• University of California, Riverside

• University of California, San Diego

• University of California, Santa Barbara

• University of California, Santa Cruz

• University of Central Florida

• University of Chicago

• University of Colorado Boulder

• University of Connecticut

• University of Delaware

• University of Florida

• University of Georgia

• University of Houston

• University of Illinois at Urbana–Champaign

• University of Iowa

• University of Kansas

• University of Kentucky

• University of Louisville

• University of Maryland, College Park

• University of Massachusetts Amherst

• University of Michigan

• University of Minnesota

• University of Missouri

• University of Nebraska–Lincoln

• University of New Mexico

• University of North Texas

• University of Oregon

• University of Pennsylvania

• University of Rochester

• University of Southern California

• University of South Carolina

• University of Tennessee

• University of Texas at Arlington

• University of Texas at Austin

• University of Utah

• University of Wisconsin–Madison

• Washington University in St. Louis

• West Virginia University

• Vanderbilt University

• Yale University

APPENDIX B: DESCRIPTORS AND CHARACTERISTICS (SALIENCE PERCENTAGE IN PARENTHESIS)