The study reported on in this article represents an attempt to assess the ethicality of undisclosed ghost blogging. There has been considerable debate on the blogs of public relations practitioners about whether undisclosed ghost blogging is ethical, including a recent survey of practitioners where most respondents expressed approval of undisclosed ghost blogging, provided that the stated author provides the content ideas and gives content approval.
Tiffany Derville Gallicano
University of Oregon
Yoon Y. Cho
Pusan National University
Thomas H. Bivins
University of Oregon
School of Journalism and Communication
1275 University of Oregon
Eugene, OR 97403-1275
The authors thank Toby Hopp for helpful feedback as well as the Arthur W. Page Center and the University of Oregon for funded this study. Correspondence concerning this research should be addressed to Professor Tiffany Gallicano, School of Journalism and Communication, 1275 University of Oregon, Eugene, OR 97403-1275, Telephone (+1) 541-346-2035, firstname.lastname@example.org.
Abstract: Ghost blogging occurs when a communications professional writes a blog that is attributed to someone else, such as a CEO, politician, or executive director. There has been considerable debate on PR professionals’ blogs about whether undisclosed ghost blogging is ethical (Gallicano, Brett, & Hopp, 2013). In a recent survey of public relations professionals, most respondents expressed approval of undisclosed ghost blogging, provided that the stated author provides the content ideas and gives content approval (Gallicano et al., 2013). To assess the ethicality of undisclosed ghost blogging, however, a survey of readers is needed to identify whether readers are being deceived. Consequently, this study reports the results of surveys of 507 readers of corporate blogs, 510 readers of politicians’ blogs, and 501 readers of nonprofit blogs. This study also investigates whether readers believe undisclosed ghost blogging is permissible. Wheeler and Gleason’s (1994) qualified expectation of reality is adapted as a theory for public relations, which can be used to assess whether disclosure is necessary.
CEOs, nonprofit executive directors, and politicians can use blogs to cultivate relationships, frame issues, and build thought leadership. However, blogs are time consuming and require strong writing skills.
There has been an ongoing debate in the public relations community regarding whether ghost blogging is the answer to the challenge of having executives and politicians gain the benefits of blogging without costing too much time (Gallicano, Brett, & Hopp, 2013). Communication through social media is expected to be personal, authentic, and open. However, ghostwriting is a commonly accepted practice in traditional contexts such as speechwriting.
This study explores a form of ghost blogging in which the executive or politician provides the ideas for the content of the blog posts and edits each blog post prior to publication. This same context was explored in a recent survey of public relations professionals, which found that most respondents found these conditions for ghost blogging to be permissible, although a vocal minority disagreed and the data were not generalizable (see Gallicano, Brett & Hopp, 2013). In the current study, 507 readers of corporate blogs, 510 readers of politicians’ blogs, and 501 readers of nonprofit blogs were asked whether they expected this level of blogging assistance from communications professionals and the extent to which they found this practice to be permissible.
We share our interpretations of the data with caution, noting that the data can be interpreted to support different conclusions, depending on views regarding the amount of support or opposition that is required to justify or oppose ghost blogging.
We concluded that there is support for the ethicality of having a professional write blog posts based on the stated author’s ideas, with the condition that the stated author edits each blog post, because most readers expected this to happen in the corporate and political contexts, so they were not deceived by this practice. However, enough readers in the corporate and political contexts objected to the practice, so while we believe that deception does not generally result from the form of ghost blogging we studied, we do not endorse it as an effective way to cultivate relationships.
Not enough blog readers expected ghost blogging to occur in a nonprofit context, so we decided that this practice was unethical for nonprofit executives due to the deception that would result. Not surprisingly, nonprofit blog readers also tended to object to this practice. Recommendations based on these findings are included in the conclusion of the study.
What Do Blog Readers Think? A Survey to Assess Ghost Blogging and Commenting
There is a vibrant debate in the public relations community regarding guidelines for ghost blogging (see Defren, 2010; Fleet, 2009; Harte, 2009; Sledzik, 2009). Advocates for undisclosed ghost blogging offer the same reasons that others have used for ghostwriting speeches. First, many executives and politicians do not have the time or skills, and authenticity can be accomplished by having them share the main ideas and provide final approval (see Gallicano et al., 2013; Sledzik, 2009; Subveri, 2010, for ghost blogging and see Auer, 1984; Einhorn, 1981; Riley & Brown, 1996, for speechwriting). Second, an organizational blog or speech is really about the organization, writers often possess as much core knowledge about the topic as the stated author, and an executive or politician should use writers just as he or she uses other specialists for running the organization (see Defren, 2010; Gallicano et al., 2013; Nerad, 2010, for ghost blogging and see Einhorn, 1991, for speechwriting). In addition, audiences are sophisticated and probably expect that an executive or politician has had assistance with the blog (see Gallicano et al., 2013; Nerad, 2010, for ghost blogging and see Riley & Brown, 1996, for speechwriting).
Among the defenders of undisclosed ghost blogging is the Public Relations Society of America (PRSA). Although the code of ethics does not specify whether organizations’ or politicians’ ghostwritten blogs should include disclosure statements (similar to the ethical code of the Word of Mouth Marketing Association), PRSA has provided tacit endorsement through two articles in Tactics, one of its publications. In the first article, the author explained that ghost blogging is ethical, as long as the stated author is involved in the creation of the post, the stated author gives final approval of the post, and the post does not mislead readers (Subveri, 2010). In the second article, the author described the importance of getting feedback from the stated author when engaging in ghost blogging and other ghostwriting practices (O’Brien, 2011).
Critics of undisclosed ghost blogging, however, explain that the rules for social media differ from the rules for traditional tactics (Gallicano et al., 2013; Fleet, 2009; Harte, 2009). Blogging is used to build personal relationships. As Fleet (2009) noted, “When did outsourcing your relationships become okay?” (para. 5). The argument that social media is a relationship-building medium is supported by scholarly research. For example, interactivity—a key component of social media—seems to have a significant impact on relationship building (Bortree & Seltzer, 2009; Kent & Taylor, 2002; Seltzer & Mitrook, 2007). Another problem is that ghost blogging can lead to inflated assessments of an executive’s skills and the organization’s future (Terilli & Arnorsdottir, 2008). In addition, if audiences do not expect ghost blogging, then the practice is deceptive (see Gallicano et al., 2013; Fleet, 2009; Harte, 2009, for ghost blogging and Bormann, 1956, for speechwriting). Also focusing on audience opinion, Arthur Page (2011) in 1927 stated that public relations “ought to act all the time from the public point of view, even when that seems in conflict with the operating point of view. It ought to bring to the management at all times what it thinks the public is going to feel about a thing” (p. 22).
There is theoretical support for the position that the ethicality of a practice hinges on the audience’s expectations. The qualified expectation of reality theory (QER) has been used to assess the ethics of photograph alteration (Wheeler & Gleason, 1994; Wheeler, 2002). In this study, we use the theory to explore the ethics of undisclosed ghost blogging under the conditions that the content comes from the executive and the executive provides content approval.
This study accomplishes three goals. First, it extends the scope and utility of the QER theory by conceptualizing it and using it to explore the ethicality of social media decisions. Second, it provides results regarding the ethicality of undisclosed ghost blogging through the application of the theory. Even if a practice is ethical, however, it is not necessarily an effective strategy for cultivating relationships, so our third goal involves discovering the permissibility of undisclosed ghost blogging from the perspectives of the readers of corporate blogs, politicians’ blogs, and nonprofit blogs.
A survey of 479 practitioners reported that 57 percent thought that blogs were either somewhat important or very important to their organization’s public relations efforts (Wright & Hinson, 2011)—and for good reason. Organizational blogs can be a valuable tactic for relationship building (e.g., Kelleher & Miller, 2006; Seltzer & Mitrook, 2007; Teich, 2008; Terilli & Arnorsdottir, 2008; Yang & Kang, 2009). In addition, organizations can use blogs to build an organization’s image (Gilpin, 2010), frame issues (Gilpin, 2010; Terilli & Arnorsdottir, 2008), generate media coverage (Gallicano, 2014; Waters, Tindall & Morton, 2010), and show thought leadership (Terilli & Arnorsdottir, 2008), which can result in new business and client retention (Porter, Sweetser Trammell, Chung, & Kim, 2007). Given the value attached to organizational blogs, it is important to explore the potential ethical issues related to how they are written and by whom.
Qualified Expectation of Reality Theory
Wheeler and Gleason (1994) focused on whether an audience is deceived to establish a theory for assessing the ethics of photograph alteration (also see Wheeler, 2002). The qualified expectation of reality theory is based on the idea of an audience’s expectations of the degree to which a communication artifact (such as a photograph or in this case, a blog post or comment) represents reality. Because the theory is receiver-focused, communicators can consider whether to disclose information by focusing on audience expectations (i.e., in this case, whether the audience expects public relations assistance with an executive’s or politician’s blog or whether an audience is deceived by the byline). The focus on audience expectations allows the QER test to be modified to determine the ethicality of communication of all sorts, such as audio editing (Huntsberger, 2008), public relations, and advertising (Bivins, 2009). The theory includes four tests.
The first two tests involve the extent to which people identify the fictitious nature of the content. We categorize undisclosed ghost blogging and ghost commenting as fiction because the stated author is not the real author, even if the ideas and final approval come from the stated author. With the first test, the producer (as applied to this study, the public relations practitioner) reflects upon whether the fictitious content is immediately obvious. If reasonable people identify the fiction, then no one is fooled and the act is ethical by default.
With the second test, the producer considers whether he or she thinks that people can identify the fiction based on the context. For example, the context of tabloids would suggest photo alteration. In our scenario, if reasonable people expect that a politician or executive would have assistance with all written public communication, deceit does not occur. Due to practitioners’ arguments that expectations are different for social media than they are for work such as speechwriting (see Fleet, 2009; Gallicano et al., 2013; Harte, 2009), discovering blog readers’ expectations is an important aspect of this study.
The third and fourth tests, while not immediately relevant to ethicality, do provide insight into other aspects that are relevant to ghost blogging and ghost commenting. Based on QER, the third test may be used if the communication artifact fails the first two tests. For the third test, the producer checks to see if appropriate disclosure is provided so that no reasonable person is deceived (Wheeler & Gleason, 1994). For example, if audiences do not expect ghost blogging, then ghostwritten blogs are ethically required to include disclosure statements unless the ghostwriting assistance is so minimal that it passes the fourth test. The fourth test permits the fiction as long as the fiction is irrelevant to the essence of the communication artifact (in this case, the blog or comment). For example, if an executive or politician drafts a blog post and a public relations practitioner fixes the comma use, this editing would be considered irrelevant to the essence of the artifact and it would be considered ethical.
Whereas the first focus of our study is ethics, the second focus is about whether blog readers find ghost blogging and ghost commenting permissible, which is a matter of relationship management rather than ethics (see Wheeler, 2002). If readers are not deceived, then the choice to have a public relations practitioner draft a blog for an executive or politician without disclosure is ethical because no reasonable person is fooled. Nevertheless, if readers identify the practice but disapprove of it, the practice would not be effective for building lasting relationships.
Ghost blogging. The practice of ghost blogging refers to “writing blog posts on behalf of someone else who is stated as the author” (Gallicano et al., 2013, p. 2). Ghost commenting can be defined as the practice of writing comments online on behalf of someone else who is stated as the author, and it can exist on the stated author’s social media sites or on others’ sites. We propose, then, that the central issue concerning ghost blogging is one of transparency.
This study is focused on politicians’ blogs and organizational blogs (i.e., blogs that officially represent the organization), as opposed to personal, organization-aligned blogs in which the expressed opinions belong to the author and do not necessarily reflect the organization’s positions, even though the organization is mentioned on the blog (Defren, 2010; Gallicano et al., 2013). Ghost blogging can occur with blogs that contain disclosure statements about a public relations practitioner’s assistance and with blogs that do not contain disclosure (Gallicano et al., 2013; Fleet, 2009).
Preliminary research suggests that ghost blogging without disclosure could be common. A content analysis of CEO blogs found that five blogs contained disclosure statements about assistance from other writers or editors while 40 blogs did not (Terilli & Arnorsdottir, 2008). Of course, the sample size was small, and it is possible that CEOs actually authored the blogs that did not have disclosure statements. A survey of 291 public relations practitioners revealed that about 41% of the participants had at least one executive blog, and a little more than half of these respondents indicated that their executive blogs were ghostwritten (Gallicano et al., 2013). From this study, 71% of practitioners thought that drafting blog posts for an executive without public disclosure is ethically acceptable, as long as the ideas come from the executive and the executive provides final approval.
The results of Gallicano et al.’s (2013) survey could be surprising, given public relations scholars’ emphasis on honesty and disclosure: “Being transparent in public relations (or any form of public communication) means that both your identity and your motivation are apparent to those you are trying to persuade” (Bivins, 2009 p. 180; also see Sweetser, 2010). The majority of case studies about unethical online practices involve concealing information, resulting in crises when the online community exposes the truth (Martin & Smith, 2008).
In this climate, there is a movement among practitioners toward radical transparency (Beal & Strauss, 2008). In public relations, the concept of radical transparency is “a philosophy for doing business that refers to prioritizing transparency above all other competing values, with the exception of disclosing information that violates regulations or ethical principles” (Gallicano et al., 2013, p. 8). The idea is that radical transparency can lead to increased trust, which can significantly improve relationship building, especially given that corporate blogs are trusted by only 16% of the people who read them who do not have blogs themselves (Bernoff, 2008). Radical transparency is likely to result in a competitive advantage only when an organization acts ethically and is competitive in the marketplace (Gallicano et al., 2013). Otherwise, radical transparency could result in people avoiding an organization because of its unethical practices or because it is a poor investment. Radical transparency has gained momentum in the public relations community. For example, Richard Edelman (2011), Ketchum president Rob Flaherty (2011), and Bob Pickard, president and CEO of Burson-Marsteller Asia Pacific, have emphasized radical transparency (de Vera, 2011).
Practitioners who want to adopt the philosophy of radical transparency can consider alternatives to undisclosed organizational ghost blogging. Instead of having a CEO as an author, an organization could have a different employee serve as the stated author who has time to write the blog, or even a group of employees could jointly author the blog (Beal & Strauss, 2008; Fleet, 2009). A second option is for the executive to use a social media channel that he or she has time for rather than a blog, which tends to be especially time consuming (Fleet, 2009). Another option is disclosure. Holtz (2011) provided an example:
Welcome to my blog. Several times each week, I articulate my thoughts to Mary Jones, who runs communications for the company, and she posts them here ensuring that I make the points I want to make. But rest assured, while Mary makes me sound better, the messages you read are mine; they come from my heart and I read all the comments myself. (para. 3)
Practitioners want to learn more about how to “engage in a meaningful and transparent manner” online, and they want to know more about social media policies that minimize risk (DiStaso, McCorkindale, & Wright, 2011, p. 327). This study helps illuminate this area by showing how transparent practitioners need to be with regard to ghost blogging and commenting. To discover whether deceit results from ghost blogging, we explored the following question:
RQ 1: To what extent do blog readers expect companies, politicians, and nonprofits to engage in ghost blogging?
Considering that 71% of the practitioners in a recent study (Gallicano et al., 2013) indicated that it is permissible to engage in undisclosed ghost blogging we wanted to see how practitioners’ opinions about acceptability compared with readers’ opinions, so we asked a second research question:
RQ 2: To what extent do blog readers think it is permissible for companies, politicians, and nonprofits to engage in undisclosed ghost blogging, provided that the ideas come from the stated author and the stated author approves the message?
To learn whether audiences are being deceived by undisclosed ghost commenting, we pursued the following question:
RQ 3: To what extent do blog readers expect companies, politicians, and nonprofits to have someone write comments in reply to readers’ comments on behalf of the stated author without disclosure?
Notably, just over half (56%) of the practitioners in a recent study think that that it is permissible for an executive to have a staff member write comments in reply to readers’ comments on the executive’s blog without disclosure (Gallicano et al., 2013). To see how this compared with readers’ opinions, we presented the following question:
RQ 4: To what extent do blog readers think it is okay for companies, politicians, and nonprofits to have someone write comments in reply to readers’ comments on behalf of the stated author without disclosure, provided that the ideas come from the stated author and the stated author approves the message?
To explore this area further by looking at ghost commenting on other people’s blogs, we explored the following question:
RQ 5: To what extent do blog readers expect companies, politicians, and nonprofits to have someone write comments on other people’s blogs on behalf of the stated author without disclosure?
Less than half (42%) of the practitioners in a recent study thought that it was permissible to engage in ghost commenting on other people’s blogs (Gallicano et al., 2013). To discover how this compares with readers’ views, we asked the following question:
RQ 6: To what extent do blog readers think it is permissible for companies, politicians, and nonprofits to have someone write comments on other people’s blogs on behalf of the stated author without disclosure, provided that the ideas come from the stated author and the stated author approves the message?
We partnered with Qualtrics to conduct three nationwide surveys. Partnering with a research company enabled us to recruit readers of corporate blogs, politicians’ blogs, and nonprofit blogs from a population that was representative of the U.S. Census. Anyone who read at least one corporate blog per month was qualified to participate in the survey of corporate blogs, and we applied this rule through a screening question. Parallel rules and screening questions were in place for people who participated in the surveys about politicians’ blogs and the nonprofit blogs. Collecting data about the contexts separately enabled a comparison across categories.
Respondents had volunteered with the research company to participate in surveys in exchange for compensation. Participants were compensated through a cash value reward that was credited to their account with the survey company. They could redeem the money through a gift card to an organization such as Amazon. The incentive the research company offered to participants was $2. The research company has several security measures in place for its panel of participants, such as preventing multiple registrations from the same computer. It also prevents participants from completing more than one survey every 10 days.
A total of 1,518 blog readers participated in the study. Specifically, 507 participated in the survey of corporate blogs, 510 participated in the questions about politicians’ blogs, and 501 answered the questions about nonprofit organizations’ blogs. The response rate was 14.1%. Studies with similar or lower response rates include 5.6% for Kang (2010); 11.9% response rate for Porter, Sweetser, and Chung (2009); and 14% for Porter, Sweetser Trammell, Chung, and Kim (2007).
The demographic characteristics of participants in the corporate, political, and nonprofit surveys were similar, so we report them here as an average of the three groups. Excluding people who did not disclose their information and with rounding (which is why the numbers do not add up to 100%), the surveys together included approximately 43% men and 56% women. For race, the surveys included 76% Caucasians, 8% African-Americans, 5% Asians/Pacific Islanders, 2% Native Americans, and 2% multiple-race people. Of the respondents, 6% were Hispanic. Age was measured by using dates commonly applied in studies for categorizing generations (e.g., Twenge, Campbell, Hoffman, & Lance, 2010). In this study, approximately 4% represented the Silent Generation (born 1925-1945), 30% were Baby Boomers (born 1946-1964), 38% came from Generation X (born 1965-1981), and 27% were Millennials (born in 1982 or later). In terms of income, 21% earned $29,000 or less, 25% earned $30,000-$49,000, 22% earned $50,000-$69,000, 11% earned $70,000-$89,000, 14% earned $90,000 or more, and several people declined to answer.
To explore blog readers’ responses in three contexts (corporate, political, and nonprofit), we created three versions of a survey. Participants responded to a screening question about whether they read a particular type of blog at least once a month. The particular type of blog was a company’s blog, politician’s blog, or nonprofit’s blog, depending on the survey. If qualified, participants were instructed to answer the main questions about their expectations and acceptance or rejection of ghost blogging. The corporate version of the main questions can be seen in Tables 1, 3, and 4, and an example of how the questions were adjusted to fit to the nonprofit and political surveys can be seen in the next paragraph. The survey concluded with questions about demographics.
Participants rated their agreement or disagreement with statements on a 5-point Likert scale. For example, we measured expectations about ghost blogging on a politician’s blog to assess whether deceit was occurring by examining the extent to which respondents agreed or disagreed with statements such as, “I think it’s common if a politician is listed as the author of a blog, even though it’s really written by someone else – as long as the ideas come from the politician and this person approves the message.” An example of a question about the permissibility of ghost blogging in the nonprofit context was, “I think it’s okay if a nonprofit’s president or executive director is listed as the author of a blog, even though it’s really written by someone else – as long as the ideas come from the president or executive director and this person approves the message.” Parallel statements were in place for participants who responded to the other survey versions.
Single-item measures were used for each research question. Although we believe the items were basic enough to justify this compromise, our use of single-item measures is a limitation. Access to a random sample of people who read corporate blogs, nonprofit blogs and politicians’ blogs was expensive, and the length of the survey affected the cost. We believe the tradeoff is justified; however, we encourage the use of additional research that uses multiple item measurements.
We used SPSS 18.0 as the statistical package to analyze the data for the research questions. To determine whether readers’ opinions differed about ghost blogging expectations and acceptance based on the type of blog, we used a multivariate analysis of variance (MANOVA). Then, a one-way ANOVA and Tukey’s post hoc test were run to see mean differences among groups of respondents on each dependent variable. Finally, an additional one-way ANOVA was conducted to see if gender is related to readers’ expectations or acceptance of ghost blogging and ghost commenting.
Before reporting the results of each research question, a one-way MANOVA was performed to determine whether the three groups of respondents have different opinions regarding expectations and acceptance of ghost blogging and commenting. Using Wilk’s test of multivariate significance, three groups of respondents showed significantly different opinions regarding ghost blogging, ghost commenting, and ghost commenting on other blogs (Wilk’s Λ = .95, F (12, 3020) = 6.47, p < .000, η2 = .03).
RQ 1: Expectations of Ghost Blogging
As shown in Table 1, 57.2% of corporate blog readers agreed or strongly agreed that they think it is common for a company to have a blog that lists the president or CEO as the author, even though it is really written by someone else, and 52.1% of the readers of politicians’ blogs agreed or strongly agreed with the same question about politicians. Less than 20% disagreed or strongly disagreed that they expected ghost blogging in either the corporate or politician survey, 26% of corporate blog readers were neutral, and 28% of the readers of politicians’ blogs were neutral. Support was lower for the nonprofit condition, with only 34.2% agreeing or strongly agreeing that they expected ghost blogging by nonprofit executives, 27% disagreeing or strongly disagreeing, and 38.9% who were neutral. A univariate ANOVA on expectations about ghost blogging revealed statistically significant mean differences among three groups of respondents, F (2, 1515) = 30.16, p < .000. Tukey’s post hoc tests indicated that corporate blog readers were the most likely to expect ghost blogging (Mc = 3.59, SDc = 1.12), followed by readers of politicians’ blogs (Mp = 3.43, SDp = 1.17) and readers of nonprofit blogs (Mn = 3.05, SDn = 1.11, see Table 2).
RQ 2: Permissibility of Ghost Blogging
Table 1 also depicts the extent to which blog readers agreed that it is permissible for the stated author to engage in undisclosed ghost blogging, provided that the ideas come from the stated author and the stated author approves the message. Agreement with this practice was substantially lower than people’s expectations for its occurrence. Although respondents expected ghost blogging to exist in corporate, political, and nonprofit contexts (Maverage = 3.36), they did not think it was an acceptable practice (Maverage = 2.84).
About 39.6% of corporate blog readers agreed or strongly agreed that ghost blogging is acceptable if the ideas come from the stated author and the stated author gives content approval, 38% disagreed or strongly disagreed with this statement, and 22.3% were neutral. About 33.9% of the people who read politicians’ blogs agreed or strongly agreed that ghost blogging is permissible, 44.5% disagreed or strongly disagreed, and 21.6% were neutral. The disagreement rate among nonprofit blog readers was the highest, with 46.6% disagreeing or strongly disagreeing with ghost blogging, while only 35.5% of nonprofit blog readers approved of the practice, and 18% were neutral. A univariate ANOVA confirmed that the three groups of respondents showed different opinions regarding acceptance of ghost blogging, F (2, 1515) = 3.07, p < .05. Tukey’s post hoc tests revealed that corporate blog readers were the most likely group to accept ghost blogging: (Mc = 2.95, SDc = 1.30), ghost blogging was least acceptable to nonprofit blog readers (Mn = 2.75, SDn = 1.34), and readers of politicians’ blogs fell in the middle of these two groups in terms of ghost blogging acceptability (Mp= 2.81, SDp = 1.36, see Table 2).
RQ 3: Expectations of Comment Help on Author’s Blog
As seen in Table 3, 64.1% of corporate blog readers agreed or strongly agreed that it is common for the stated author to have a staff member write comments in reply to readers’ comments on the stated author’s blog without a disclosure statement that a staff member is responding on behalf of the stated author. Only 14.6% of corporate blog readers disagreed or strongly disagreed that this practice is common, and 21.3% were neutral. Similarly, 64.8% of the respondents who read politicians’ blogs agreed or strongly agreed that it is common, while only 16.5% disagreed or strongly disagreed that ghost commenting on the stated author’s blog is common, and 18.8% were neutral. Expectations were lower for nonprofit blog readers than for the other two groups, with 52.9% agreeing or strongly agreeing that the practice is common, 23.2% disagreeing or strongly disagreeing, and 24% who were neutral. A one-way ANOVA on expectations about ghost commenting confirmed significant differences among three groups of respondents, F (2, 1515) = 14.51, p < .000. Tukey’s post hoc tests confirmed that corporate blog readers (Mc = 3.73, SDc = 1.06) were the group that was most likely to expect that it is common for someone to ghostwrite a stated author’s comments without disclosure, followed by readers of politicians’ blogs (Mp= 3.69, SDp = 1.16) and then readers of nonprofit blogs (Mn = 3.38, SDn = 1.19, see Table 2).
RQ 4: Permissibility of Comment Help on Author’s Blog
Whereas the third research question examined ghost commenting expectations, the fourth question explored the extent to which blog readers agree that it is permissible to engage in ghost commenting without disclosure on the stated author’s blog. A one-way ANOVA test confirmed that there was no significant difference among these groups regarding acceptance of ghost commenting (p = .08).
As shown in Table 3, the differences in opinions among the three groups were not significant. Less than half of the blog readers approved of ghost commenting, with 40.1% agreeing or strongly agreeing that the practice is permissible in the corporate survey, 37.1% agreeing or strongly agreeing with the practice in the politician survey, and 35.5% agreeing or strongly agreeing with the practice in the nonprofit survey. Meanwhile, 35.5% of corporate blog readers disagreed or strongly disagreed that ghost commenting is acceptable, 39.4% of the readers of politicians’ blogs disagreed or strongly disagreed, and 43.6% of nonprofit blog readers disagreed or strongly disagreed. In the corporate survey, 24.5% were neutral; in the politician survey, 23.5% were neutral; and in the nonprofit survey, 21.2% were neutral.
RQ 5: Expectations of Comment Help on Other Blogs
As depicted in Table 4, 51.9% of corporate blog readers agreed or strongly agreed that it is common for the stated author to have a staff member write comments in reply to other people’s blogs without a disclosure statement that it is a staff member commenting on behalf of the stated author, 20.1% of corporate blog readers disagreed or strongly disagreed, and 28% of corporate blog readers were neutral. Of the readers of politicians’ blogs, 50.6% agreed or strongly agreed that ghost commenting on other people’s blogs is common, whereas 23.9% disagreed or strongly disagreed, and 25.5% were neutral. For the nonprofit survey, 43.9% of readers agreed or strongly agreed, 28.2% disagreed or strongly disagreed, and 27.9% were neutral. A one-way ANOVA test showed that there was a significant difference among the three groups of respondents regarding expectations about ghost commenting on other blogs, F (2, 1515) = 8.24, p < .000. Tukey’s post hoc tests confirmed that corporate blog readers were most likely to expect ghost commenting on other people’s blogs (Mc = 3.46, SDc = 1.15), followed by readers of politicians’ blogs (Mp = 3.36, SDp = 1.22) and then readers of nonprofit blogs (Mn = 3.16, SDn = 1.19, see Table 2).
RQ 6: Permissibility of Comment Help on Other Blogs
Table 4 also shows the extent to which blog readers think that ghost commenting on other people’s blogs without disclosure is an acceptable practice, provided that the ideas come from the stated author and the stated author approves the message. A one-way ANOVA test confirmed that the three groups of respondents did not show different opinions about this practice (p = .16). Roughly a third of the readers found the practice to be acceptable, with 33% agreeing or strongly agreeing that it is an acceptable practice in the corporate survey, 30.8% agreeing or strongly agreeing in the nonprofit survey, and 27.7% agreeing or strongly agreeing in the politician survey. The neutral category was selected by 23.3% of corporate blog readers, 18.8% of nonprofit blog readers, and 23.7% of the readers of politicians’ blogs. Meanwhile, 43.8% of corporate readers disagreed or strongly disagreed with ghost commenting on others’ blogs, 48.7% of the readers of politicians’ blogs disagreed or strongly disagreed, and 50.4% of nonprofit readers disagreed or strongly disagreed.
Demographics and Correlations
An analysis of demographic correlations revealed only one major relationship of importance. As shown in Table 5, up to 50% of the female respondents either strongly disagreed or disagreed with ghost blogging and ghost commenting, as compared with the male respondents (less than 40% of the men either strongly disagreed or disagreed). A one-way ANOVA test confirmed that women were less likely than men to accept ghost blogging (F (1, 1510) = 13.26, p < .000), ghost commenting on the stated authors’ own blogs (F (1, 1510) = 5.46, p < .05), and ghost commenting on other people’s blogs (F (1, 1510) = 6.48, p < .05). Regarding expectations, women were less likely than men to expect ghost commenting on others’ blogs (F (1, 1515) = 5.16, p < .05, see Table 6).
Based on the QER test (Wheeler & Gleason, 1994), we interpret the results of the first research question as providing ethical justification for corporate executives and politicians to engage in ghost blogging, considering that more than half expect the practice and less than 20% disagree with it. However, we do not think that the results are positive enough in the nonprofit survey to provide ethical justification for ghost blogging, with a little over a third expecting it, a little less than a third not expecting it, and the rest who are not sure either way.
We explored the permissibility of ghost blogging through the second research question. Participants were more likely to expect ghost blogging than they were to find the practice permissible, even with the ideas for the content coming from the stated author and with the stated author providing content approval. The agreement and disagreement rate among corporate blog readers was nearly equal (at nearly 40% on each side), and more people disapproved of the practice than people who approved of it in the other two groups. With disapproval at nearly 40 percent in the corporate group and in the mid-40s in the other two groups, this is a substantial enough number in our view to classify undisclosed organizational ghost blogging as a poor way to cultivate positive relationships with readers in all three contexts. This is an important finding, considering that 71% of practitioners in a recent study expressed the view that this practice is permissible, 20.7% disapproved of it, and the rest were neutral (and there were not any differences in opinion by the practitioner’s work setting, such as nonprofit or corporate; Gallicano et al., 2013). The mean on this question was a 2.83 among blog readers in this study, and it was a 3.65 among practitioners in a recent study (Gallicano et al., 2013).
The third research question focuses on expectations regarding ghost commenting on the stated author’s blog. Similar to the first research question, corporate blog readers were significantly more likely to expect ghost commenting on the stated author’s blog, followed by readers of politicians’ blogs and then readers of nonprofit blogs. Despite the differences, the disagreement numbers were low enough in every context that ghost commenting passes the qualified expectation of reality test from our perspective. What is particularly interesting is that an average of 60.6% of participants across the three groups expected ghost commenting, whereas an average of 47.8% of participants across the three groups expected ghost blogging. The disapproval rate for ghost blogging was also slightly higher than it was for ghost commenting across the three groups. It seems that readers are more likely to expect the stated author to write the blog post and delegate the commenting than the other way around.
In research question four, we explored the permissibility of ghost commenting on the stated author’s blog from the perspectives of blog readers. There was not a significant difference in permissibility among the three groups. Across the three groups, an average of 37.5% of the participants thought the practice was acceptable (as compared with 56.3% of the practitioners from Gallicano et al.’s 2013 study who approved) and 39.5% disapproved (as compared with 35.4% of practitioners from Gallicano et al.’s study who disapproved). With the disapproval rate at nearly 40%, our interpretation is that ghost commenting on the stated author’s blog is not effective for cultivating long-term relationships. An interesting finding is that practitioners’ approval of ghost blogging was 71.1%, and this approval dropped to 56.3% when asked about ghost commenting on the stated author’s blog. However, an average of 36.3% of blog readers across the three groups approved of ghost blogging, and this number inched up to 37.5% for ghost commenting on the stated author’s blog. Thus, a significant number of practitioners in Gallicano et al.’s study made a distinction between the permissibility of ghost blogging and ghost commenting; however, the blog readers in our study did not.
Research question five explored ghost commenting by the stated author on other people’s blogs. Corporate blog readers were significantly more likely to expect this practice, followed by readers of politicians’ blogs and then readers of nonprofit blogs. With only 20.1%-28.2% of readers not expecting this practice in the three groups, we believe that this practice passes the qualified expectation of reality test. Not surprisingly, expectations were higher for ghost commenting on the stated author’s own blog than they were for ghost commenting on other people’s blogs, with an average of 60.6% of readers expecting ghost commenting on the stated author’s blog and an average of 48.8% of readers expecting ghost commenting on other people’s blogs.
The three groups did not significantly differ in their acceptance of ghost commenting on other people’s blogs for the sixth research question. With an average of 30.5% of the readers in the three groups thinking the practice is acceptable (compared with 42.6% of practitioners who approved of it in Gallicano et al.’s 2013 study) and 47.6% disapproving of the practice (as compared with 44% of practitioners who disapproved of it in Gallicano et al.’s study), we believe this practice is not advisable for cultivating long-term relationships. In terms of co-orientation, public relations practitioners and blog readers appear to be the closest in their response to this question, mostly because practitioners’ approval of this particular practice substantially plummeted, as compared with the other practices.
Corporate blog readers and readers of politicians’ blogs were nearly always more likely to expect ghost blogging and ghost commenting and were more likely to agree that these practices are acceptable than nonprofit blog readers. One possible explanation is that readers could expect nonprofits to run on a shoestring budget, thinking that the executive should do this work himself or herself, in part due to a deep commitment to the organization’s mission. In addition, women were consistently less likely than men to accept ghost blogging and ghost commenting across the three groups. This could be the case because women might generally be more invested than men are in the relationship-building function of blogs and blog comments, and arguably, having a ghostwriter limits the relationship building appeal of social media.
In this study, we extended the scope and utility of the qualified expectation of reality theory by adapting the conceptualization to an area of social media ethics and using it to explore ghost blogging. Based on the data, we believe that ghost blogging and ghost commenting pass the qualified expectation of reality test for companies and politicians. We did not think that enough people expected ghost blogging in the nonprofit context to justify ethical endorsement; however, ghost commenting on other people’s blogs and the stated author’s own blog was widely expected in the nonprofit condition and passed the qualified expectation of reality test in our view. This was consistent with our data, which suggested that readers are more likely to expect ghost commenting than ghost blogging, which is interesting in comparison to the higher support from practitioners in Gallicano et al.’s (2013) study for ghost blogging than ghost commenting (practitioners’ higher support for ghost blogging is possibly due to the interaction involved in ghost commenting).
Although we found ethical justification for ghost blogging and ghost commenting in all contexts except the nonprofit ghost blogging context, we also discovered that permissibility of these practices was not high enough to endorse either of them as long-term relationship strategies for any of the three contexts we examined. Consequently, if an executive or politician is going to have a ghostwritten blog, it should contain a disclosure statement. In addition, if the executive or politician engages in ghost commenting on his or her blog, we recommend that the blog disclosure statement also include the topic of ghost commenting. We believe it is not practical to have a disclosure statement for ghost commenting on other blogs unless the disclosure is included in the comment itself, so we advise against the practice of ghost commenting on other blogs.
We expect that practitioners who approve of ghost blogging without disclosure might believe that including a disclosure statement could limit reader engagement with the blog, thus limiting the blog’s effectiveness (in the sense that readers believe that the stated author wrote it). However, if practitioners worry that readers would not find the blog to be as engaging if they knew it was ghostwritten, then practitioners’ motivation for not revealing this information is deception, which is unethical. Although some professionals might see this as a purely functional issue, it is actually a moral issue. As Plaisance (2007) points out:
Communication is based on the notion of honest exchange. This norm of forthrightness, or being ‘‘aboveboard,’’ is what is known as being transparent. And society would not be possible if we did not place a premium on the spirit of openness, or transparent behavior. (p. 191)
Moreover, our research suggests that audiences expect company presidents, CEOs, and politicians to have someone else draft blogs for them anyway, so a disclosure statement could possibly make a ghostwritten blog more credible, especially given research that suggests that most corporate blog readers do not trust corporate blogs (Bernoff, 2008). In this way, this study lends theoretical support for either embracing radical transparency by disclosing ghost blogging and ghost commenting or choosing an alternative to these practices described in our literature review. Furthermore, practitioners should not place themselves in a situation in which they would be concerned about readers finding out that the blog is not really drafted by the stated author, given the amount of disapproval of the practice we found in this study.
If we follow the model suggested by Wheeler and Gleason (1994), a practitioner’s obligation is to remove all doubt concerning the reality of the context of his or her message. The goal should be to eliminate any potential for misunderstanding. If our goal is to intentionally blur the lines between fact and fiction, we are acting unethically (Bivins, 2009).
Limitations, Future Research, and Significance
Although the study has a significant number of randomly selected participants, the results could be compromised if research volunteers fundamentally differ from blog readers who do not volunteer with research firms. The use of multiple-item measures would have strengthened the credibility of this study. Nevertheless, the design of this study has strong advantages, including the sample size and random sampling method from a U.S. Census-representative population. Due to the sample, the findings can only be generalized to U.S. blog readers.
There are several directions for future research. One direction involves discovering the opinions of readers from other countries. In addition, research is needed to discover how effective an executive ghostwritten blog is that has a disclosure note, such as the one offered by Holtz (2011), in terms of engagement and other outcomes. Research is also needed to find out whether it would be a smarter decision not to blog at all or to have the organizational blog authored by a group of employees who write all of their blog posts rather than having a ghostwritten blog with disclosure.
What could be inferred from this study is that it is better for long-term relationship building to have a ghostwritten blog with a disclosure statement than it is to have one without it, because blog readers appear to expect ghost blogging anyway and disapprove of it. Moreover, if an executive or politician writes his or her own blog, a disclosure statement saying so would appear to be good for relationship building, given the number of people who seem to expect ghostwritten blogs (especially for companies and politicians). In addition, this study shows that audience expectations and permissibility for particular practices could be a matter of context. As such, we caution against generalizing studies about audience expectations and permissibility across contexts. This study is also significant by adapting the qualified expectation of reality theory to a public relations context. Scholars could use this theory in future studies to identify the research that is needed to determine whether disclosure is necessary for the ethical practice of public relations, especially considering the expectations for heightened transparency with social media use.