A Framework for Messy Communication: A Qualitative Study of Competing Voices of Authority on Social Media

Abstract

While governments, nonprofits, and influencers differ dramatically in their resource availability for social media content creation and their duties to public safety, health, and welfare, all play a role in communication with the public at large. Governments provide for the broader public good and nonprofits and influencers have the opportunity to serve isolated communities vulnerable to social ills and health crises. We explore how these three content creator groups concurrently and independently use social media to provide critical information. We qualitatively coded 1392 posts by governments, nonprofits, and influencers on 6 social media platforms shared across the United States during the 2022 Mpox outbreak. We constructed a framework defining a public health communication progression from chaotic to controlled. The findings indicated that governments can reach minoritized communities to resolve a public health crisis by partnering with nonprofits and influencers trusted by these audiences and, most importantly, practicing flexible control over shared messaging.

Life is messy. Public administration is dynamic. Communication is ever-changing. Crisis-related communication strategies are often disorganized due to an absence of thoughtful planning (Mergel and Bretschneider 2013). Instead, public administrators concentrate on putting out fires and managing the crisis at hand. Therefore, how do public administrators use social media to distribute information and respond to constituents and stakeholders during critical moments? With the ever-evolving landscape of social media, this is an important question. Popular social media users with influence sometimes step in with attractive and concise messaging that is appealing to users (Albadri 2023) but is not always reliable or accurate (Mishra and Ashfaq 2023) to shift the public response to critical issues (Peter and Muth 2023). More people than ever obtain their news and information through social media platforms, such as Facebook, Twitter/X1, Instagram, and TikTok (Matsa 2022; Shearer 2023). Furthermore, with misinformation and disinformation easily spread (Mishra and Ashfaq 2023; Hilton and O’Leary 2018), public organizations have to provide timely and trusted information. Here, we construct a framework where government employees can work through this messiness by working with nonprofits and social media influencers similar to their for-profit counterparts (Joshi et al. 2023). These three content creator groups (i.e., government, nonprofits, and influencers) can reach out to minoritized communities, create specialized messaging, and effectively use social media.

We explore how nonprofits, governments, and influencers used social media concurrently and independently to provide information and support to communities during the Mpox2 outbreak. The Mpox outbreak is a strong example of scattered public health communication in the United States (U.S.). Though the U.S. government provided the vaccinations, each state and locality was in charge of distribution, leading to large variations in distribution and governmental communication strategies. For example, in Florida, complaints about the governor’s and surgeon general’s low prioritization of Mpox in public health messaging led nonprofits to take the lead in vaccination outreach to the gay, bisexual, pansexual, and general MSM (men who have sex with men) communities (the gay community) (Sarkissian 2022). In contrast, Massachusetts directly addressed the outbreak, and when racial disparities were identified in the distribution of vaccination in Massachusetts, the state started outreach to at-risk minority communities (McCluskey 2022).

Though Mpox was not a new disease, it exploded amongst the LGBTQIA+ community, especially among gay and bisexual males, in the summer of 2022 (Beasley 2022). The first Mpox case in the U.S. was reported in Boston in late May (Beasley 2022). By June, the European Union approved a smallpox vaccine for Mpox, which the U.S. followed in July. In August, Mpox was finally declared a public health emergency in the U.S. (The American College of Surgeons 2022). Overall, there were almost 37,000 cases of Mpox in the U.S. between 2022 and 2023 (CDC Mpox in the U.S. n.d.).

In this article, we answer the research question: how do governments, nonprofits, and influencers interact to provide critical information through social media during a public health crisis? Specifically, by analyzing where, when, and what governments, nonprofits, and influencers post, we identified three intersectoral social media relationships between the three content creator groups: (1) vacuum (i.e., an absence of governmental messaging), (2) chaos (i.e., disorganized and conflicting messaging across the groups—chaos), and flex control (i.e., accurate and coordinated information sharing across all content creator groups). We contribute to public administration social media literature by providing a methodological approach for bringing influencer messaging into the scholarly dialogue and framing their position among public actors (see Zavattaro and Brainard 2019; Feeney and Porumbescu 2021). Influencers are a part of the communication network of messy communication in public administration and provide support when a public failure leads to inaction by governments and nonprofits. Influencers fill a vacuum of information. Total governmental control over social media is another possibility in public health messaging across content creator groups, but we did not observe governmental control within this context (Feeney and Porumbescu 2021). In this study, we use the progression of communication approaches (i.e., vacuum, chaos, control, and flex control) to outline a framework for understanding the messiness that is social media communication in public administration and provide recommendations for professionals and researchers.

1. Literature Review: Cleaning up the Social Media Mess

Social media provides a space for public health authorities to respond to crises in a timely, empathetic, and action-oriented way (Miller et al. 2021; Schillinger et al. 2020). This crisis response should promote self-efficacy and knowledge of at-risk audiences to affect risk protection behavior and reduce uncertainty and emotional turmoil (Miller et al. 2021). Audiences can be narrowly targeted to achieve these goals if the content creators are savvy in their social media approach (Schillinger et al. 2020; Vos and Buckner 2016). The public turns to trusted sources for direction during emergencies regardless of the source’s status as an authority in an official capacity. Bridging reliable, action-oriented information with trusted sources opens the door for effective partnerships between public organizations and influencers.

1.1. Bridging the Public Administration Online Divide

Governments, nonprofits, and influencers differ dramatically in their resource availability for content creation and their duties to public safety, health, and welfare. Governmental content creators are responsible to the public and must demonstrate reliability and accuracy in every post. On the other hand, nonprofits devote their resources toward mission-oriented goals. In some cases, the nonprofit’s duty to its funders may override pure mission focus. Regulatory constraints on programs and activities may drive publicly funded nonprofits addressing public health, safety, and welfare to align their external messaging with their funders (Corrigan 2013; Young et al. 2020). The result may be polished messaging that is misaligned with a nonprofit’s mission. Thus, while governments and nonprofits share public health goals, their execution of public health programming and communication may appear disorganized or contradictory. Other social media users may step up to help sort out the entanglement.

Social media users have grown savvy to branded content marketed directly to them (Freberg et al. 2011; Stephen 2016). Thus, when a targeted audience, such as Gen Z, seeks an online community, a sophisticated brand can actually disengage this audience because this audience is more media literate and seeks authenticity (DeMasters et al. 2024). This is where influencers may step in as trusted allies through perceived friendships (Peter and Muth 2023; Hudders et al. 2021) to promote their followers’ safety, health, and welfare (Albadri 2023). Influencers are guided by self-defined goals that promote their “brand”. However, influencers are bound by different accountability and transparency mechanisms than public administrators (Mishra and Ashfaq 2023). They are unique to this conversion because their social media generates income through direct payment from social media companies or fees for service to for-profit entities (Mishra and Ashfaq 2023; Christin and Lu 2023). Influencers lack a formal duty to the public yet possess a value essential to effective public governance: trust (Peter and Muth 2023).

We argue that social media serves as an influential public square, particularly when the COVID-19 pandemic reduced physical access to public spaces and national politics exacerbated political divides in online discourse (Hilton and O’Leary 2018). Social media influencer-generated content for the public good provides an opportunity to assess this proposition (Saldanha et al. 2023). To better understand public, nonprofit, and influencer intersections, we will first explore their social media behavior independently.

1.2. Government

How and why governments adopt social media (Mergel and Bretschneider 2013; Feeney and Porumbescu 2021; Jones 2015) versus the content of their posts (Zavattaro et al. 2015) present two distinct lines of inquiry. We focus on the latter, with the acknowledgment that social media users can be passive consumers of information (Zavattaro and Brainard 2019) with varying levels of trust in government messaging (Hilton and O’Leary 2018; Feeney and Porumbescu 2021).

The voice of the government on social media carries formal authority, similar to face-to-face interactions (Zavattaro and Brainard 2019). Appearing messy and disorganized can defeat this governmental function. The algorithmic drivers of TikTok content can make TikTok messaging appear complicated, opaque, and intimidating to figure out, which may explain governments’ hesitancy to leverage this platform. However, Zhu et al. (2020) found TikTok an effective method for Chinese Provincial Health Committees to transmit essential COVID-19 public health information. Li et al. (2021) identified similar benefits to TikTok for audience-centered risk communication on an international scale.

Zavattaro and Brainard (2019) (p. 563) argue that social media can be used to build “public values such as collaboration, transparency, and connectivity”. Governments designate resources to fulfill legislative goals, including their online messaging. They may prioritize organizational processes or outcomes and devote their pecuniary and human resources accordingly (Young et al. 2020). Federal, state, and local governments are bound by codified accountability and transparency mechanisms, which can limit their ability to quickly transmit meaningful messages to constituents during a public health emergency (Zhu et al. 2020). However, while government content creators value accuracy and reliability, social media users seek immediacy in access to information (Zavattaro and Brainard 2019). Thus, a more nimble social media content creator can more efficiently deliver valuable information into the hands of the public (Zhu et al. 2020; Li et al. 2021).

1.3. Nonprofits

Building social media capital is the first step in connecting with stakeholders online and ultimately, establishing trust. Social media capital is the attention and engagement a content creator can leverage, and it can be measured through metrics such as followers, likes, views, and reposts. However, translating social media capital into real capital and mission fulfillment is challenging (Morgan et al. 2024; Guo and Saxton 2020a). Lovejoy and Saxton (2012) developed the nonprofit social media engagement framework to explain how nonprofits engage in this space, which is also useful in making sense of government and influencer social media activity. They found that nonprofits use social media to communicate to stakeholders in three ways: sharing information, building community, and mobilizing followers into action (Lovejoy and Saxton 2012). In a meta-analysis of studies testing Lovejoy and Saxton’s framework, Campbell and Lambright (2020) found information sharing to be nonprofits’ primary social media function. This one-way communication is typical of nonprofits and governments in public health and safety messaging, as they tend to be the information authorities.

Mobilizing action is the second most used function (Campbell and Lambright 2020) and the most important in public health messaging. With ample social media capital, stakeholder mobilization can have a snowball effect, leading to crowdfunding and a critical mass of advocacy (Guo and Saxton 2020a, 2020b). Mobilizing followers to get tested or vaccinated for viruses or increase prophylactic usage can immediately affect community spread.

Third, nonprofits use social media to build an online community. While this is the least common function for nonprofits on traditional platforms, such as Facebook and Twitter/X (Campbell and Lambright 2020), Wiley et al. (2023) found that community building is the primary function for nonprofit TikTok accounts. Examples of community building include posts such as honoring volunteers, celebrating fundraising milestones, and engaging in two-way dialogue. Guidry et al. (2014) argued that interpersonal interactions and connections, as in community-building messaging, are what transform personal change or action into social change or action. Community-building content, however, does not provide clean outcome measures, as gained from information-sharing (e.g., changes in knowledge) and action-oriented content (e.g., funds raised). Assessing the value of such an approach to communication could facilitate an increased nonprofit social media capacity.

When adopting a social media platform, nonprofits prioritize their follower, like, and share counts, or social media capital, to magnify their voice (Guo and Saxton 2020b). Understanding adoption patterns and messaging strategies is useful when examining the nonprofit sector independently from its governmental counterparts. However, nonprofits are not alone in the medical and public health arena. Social media users seek and receive messaging from all sectors and user types, especially influencers (Peter and Muth 2023). In response to isolation during the COVID-19 pandemic, social media users valued connectedness (i.e., attachment, belongingness, and relatedness) over expertise, trust, and attractiveness (Saldanha et al. 2023). This reality limits the value of siloed scholarly inquiries on government or nonprofit online behavior. A comprehensive analysis across content creator groups may reveal more about the messaging social media users view and consume.

1.4. Assessing the Role of Influencers in Public Administration

Influencers are “content creators with a trusted relationship with their audience” (Michaelsen et al. 2022, p. 9) who are often considered peers by their followers (Pöyry et al. 2022). What separates influencers from governmental and nonprofit content creators is that they typically monetize their online activities (Joshi et al. 2023; Christin and Lu 2023). These individuals leverage their first-hand knowledge, reputation, professional position, experience, content quality, and personality for social capital and pecuniary gain (Joshi et al. 2023; Morteo 2018). While governments and nonprofits can leverage similar capital, influencers have more freedom to engage with users in ways public entities cannot, and they have the capacity to shift followers’ attitudes (Joshi et al. 2023). Understanding influencers’ reach and function can provide insight into how public administrators may partner with influencers, similar to how for-profit companies do (Albadri 2023; Joshi et al. 2023). For example, @nycgaydad used his followers’ trust in his parenting practices to partner with Fiskar Scissors, used by his young children in their advertisements (Jose Rolon 2023). @Larray leveraged his followers’ trust in his style choices to partner with Prada and Shein for the same purpose (Simonetti 2022).

Similar to nonprofit social media capital, influencer reach can be assessed through popularity factors, such as an overall footprint, engagement, and frequency of posts (Morteo 2018; Arora et al. 2019; Booth and Matic 2011). For instance, follower counts provide a simple way to sort influencers into mega-, macro-, and micro-influencers (Morteo 2018), but they tell us little about how their messages spread or are received. Thus, network centralities, indices, and algorithms can serve as stronger justifications for one’s role as an influencer (Pöyry et al. 2022; Arora et al. 2019; Booth and Matic 2011; Morteo 2017; Bakshy et al. 2011; Cha et al. 2010). Such measures could also inform the influencer’s role in information sharing during crises (Peter and Muth 2023).

Peter and Muth (2023) found that young adults tend to turn to mainstream media to receive the majority of their political information and they often look to influencers to make sense of political information they are learning about in school and online. Certain influencers who rarely discuss politics may not be seen as credible sources. Determining the context of influence (i.e., a profession or personal interest) and the source of their influence (i.e., reputation or experience) is an important step in assessing the value of an influencer’s public health messaging, particularly with outbreaks such as Mpox in Summer 2022 when there was little information available (Michaelsen et al. 2022; Morteo 2017; Cao 2022). For instance, public health influencers can share epidemiological research and reports due to their professional access to such information (Cao 2022). Morteo (2017) framed eight influencer categories: opinion leaders, experts, consumers, social media luminaries, celebrities, trendsetters, bloggers, and potential influencers. So many voices lead to an overcrowding of messages regarding critical issues. Thus, a follower’s trust in a particular influencer category will be guided by their reason for following them in the first place, such as when opinions align.

Opinion influencers can be categorized based on their expertise in one specific, narrow area or based on broader areas of knowledge (Pöyry et al. 2022). Personal beliefs and values affect the kind of information opinion leaders share. Thus, such influencers act as gatekeepers between government, nonprofits, corporations, news outlets, and their masses (Morteo 2018). Interestingly, when considering product messaging, social media users viewed company CEOs as more critical, skeptical, and difficult to impress than influencers. Instead, influencers were viewed as more likely to be sought out for advice and reassurance and more likely to give advice than CEOs were (Freberg et al. 2011). If users are equally critical of governmental messaging, that may challenge a government’s attempt to build public value via social media, as proposed by Zavattaro and Brainard (2019).

Opinion leaders present instructions and general facts, share campaign messages, show examples, and post emotion-laden pictures or texts. Powerful posts result in followers’ emotive reactions, further information disbursement, and the exchange of questions, advice, concern, and debate (Pöyry et al. 2022). Thus, in an epidemic, when governmental officials have few resources or facts to share, influencers, particularly experts and opinion leaders, could step in to maintain a dialogue on the public health concern and iron out some of the messiness of social media.

2. Research Design

We used a constructivist approach to assess and contrast the messaging of three content creator groups (governments, nonprofits, and influencers) across six platforms during the Mpox outbreak (Charmaz 2014). Because social media platforms are ever-evolving, the constructivist approach allowed us to make sense of social media posts within the context of each unique platform and content creator. For instance, during the writing of this paper, two main social media platforms, TikTok and Twitter/X, experienced significant changes. TikTok was banned by state governments (Franklin 2023)3, while CEO Elon Musk changed his newly purchased platform’s name from Twitter to X and lost half its users (Dempsey 2023). How quickly content creators adopt new platforms and their target audience composition vary across content creator groups (Mergel and Bretschneider 2013), necessitating a broad overview of multiple platforms. Thus, to make sense of this mess, we examined social media across popular platforms to explore one of the two propositions put forth by Zavattaro and Brainard (2019, pp. 571, 574):

“Reconceptualizing their social media as spaces for meaningful micro-encounters will shift the way public organizations use those spaces from direct information distribution to user-supplied content.”

User-supplied content from influencers captures these encounters. In this analysis, we used a deductive coding approach paired with a constructivist analysis to determine how and when public organizations leverage an online community’s trust in influencers to produce public good (i.e., public health) (Saldanha et al. 2023; Charmaz 2014).

2.1. Building the Dataset

The content creator was the unit of analysis, including government agencies, nonprofits, and influencers. To narrow our analysis, we focused on the Mpox public health crisis in six metro areas (Boston, Jacksonville, Miami, Portland, Reno, and St. Louis), with the goal of documenting all Mpox social media messaging in each metro area. This purposeful sample provided a diversity of metro areas, in consideration of policies affecting healthcare and LGBTQIA+ rights. For example, Boston and Portland are in states that are considered “blue” and more likely to support equality measures for gay men, while Tampa, Miami, St. Louis, and Jacksonville are in “red” states, less likely to support such rights. Reno is in a “purple” state. The Human Rights Campaign (2024) scored Massachusetts, Nevada, and Oregon very positively as “Working Toward Innovative Equality” for LGBTQ+ people and scored Florida and Missouri dismally as “High Priority to Achieve Basic Equality” for LGBTQ+ people. The public health systems in these states varied substantially, as ranked by the US News and World Report (2024): Massachusetts #1, Florida #15, Oregon #21, Nevada #30, and Missouri #41. Due to these variations in politics, public health, human rights, and healthcare access, we expected a variety of social media responses to this public health crisis.

The team of two university researchers and nine research assistants identified the social media accounts of national, state, and local public health agencies (ten agencies) and nonprofits with missions in public health or queer advocacy nationally and in each metro area (eighteen nonprofits). The team assembled a comprehensive list of governmental health agencies and nonprofits serving each metro area that would be expected to engage in public health crises and crises affecting queer communities (e.g., Departments of Health and Pride Centers). A team member would first check the organization’s website to locate accounts. The team members then checked the major social media platforms for the organization’s presence if their social media accounts were not linked from their web pages, including Facebook, Twitter/X, Instagram, TikTok, YouTube, LinkedIn, Snapchat, Flicker, and Pinterest.

The team then gathered the social media handles for all social media platforms for each entity and assessed whether they posted Mpox content. Those entities that did not post Mpox content were excluded. For instance, Florida’s county-based Departments of Health only post content provided at the state level. The Florida Department of Health did not post Mpox content. Thus, Florida’s governmental public health providers are not represented in this analysis.

Next, we identified health and queer influencers (10) (Agostino et al. 2019). We observed Mpox messaging on Twitter, Instagram, Facebook, and TikTok by searching #Monkeypox and #Mpox to identify active influencers in public health and queer issues during the time period. We excluded queer and health influencers who did not post about Mpox or whose content included nudity or sexual acts. We selected three queer influencers, three public health influencers, and four queer health influencers with the largest following across the six platforms who posted about Mpox.

We located all Mpox messaging for each content creator from June to October 2022. This timeframe captured the outbreak detection by public health authorities in June, treatment protocol guidance, vaccine development and distribution, and a period when the spread was under control (Beasley 2022). Online communication about Mpox drastically cut-off in September. Early October was coded to fully capture public conversation about the virus.

Team members were assigned one content creator at a time and met weekly for five weeks to collect data as a group.

The full dataset was composed of 1392 posts by 38 content creators, including Twitter/X (733), Instagram (267), Facebook (211), TikTok (132), YouTube (21), and blogs (20). We systematically selected these six social media platforms for uniformity across the content creators and dataset robustness. These platforms were used most consistently by the content creators (see also Albadri 2023). The data were a combination of descriptive quantitative data and text-based qualitative data stored in spreadsheet format. The team transformed the qualitative data into a quantitative format for descriptive analysis in SAS.

Approximately 50% of government, nonprofit, and influencer posts were on Twitter/X. Approximately 20% of government and nonprofit and 15% of influencer posts were on Instagram. The biggest differences in platform usage were between Facebook and TikTok. Facebook represented 27% and 26% of government and nonprofit posts, respectively, but only a negligible percentage of the influencer posts. On the other hand, while 23% of the influencer posts were on TikTok, only 2% of nonprofit posts were on TikTok, and there were no government posts recorded on TikTok. YouTube and blogs were also represented, though they comprised a negligible percentage of the posts.

2.2. Method of Analysis

2.2.1. Coding

We used textual and content analyses to capture and analyze Mpox public health messaging (Wiley et al. 2023; McBeth et al. 2012). First, attributes such as platform, posting date, and content creator identification information were collected. Other attributes included originality, whether it was part of a series of posts, and inclusion of trends or memes. Second, the team employed a shared codebook and coding instrument designed for collecting uniform data across the six platforms using guidance from Wiley et al. (2023). An abridged codebook can be found in Appendix A. Word-based, manifest content was copied into the survey tool. Latent and video content were documented as one would when taking field notes: a brief description or interpretation (McKee 2003). For instance, emotive tone and target audience required interpretation by the coder (Saldana 2022).

The content creator’s strategy for each post was coded to the social media framework (Lovejoy and Saxton 2012; Campbell and Lambright 2020; Wiley et al. 2023). The posts were coded as sharing information, building an online community, or mobilizing viewers to act. We extracted the data coded to information sharing for this analysis, which narrowed the dataset to 1320 posts. Informational posts were categorized by (1) the presence of location, instructions, and availability of Mpox vaccinations and treatment, and (2) descriptions, stories, and images of individual experiences with Mpox. For instance, while targeting queer communities, @PozRN shared his experience with Mpox through storytelling and videos of himself. The Center for Disease Control (CDC) posted infographics as updates on Mpox rates and vaccination sites targeting a general audience.

Lastly, the team drafted 1–2 sentence memos for each post, memos for each content creator’s platform usage, and memos accounting for all Mpox messaging by a single content creator. The memoing process resulted in three sets of memos to guide the analysis and to use as a theory-checking tool. Once a finding was considered, we referred to the memos to ensure an accurate dataset assessment.

2.2.2. Analysis

The first round of analysis occurred when the authors were coding the data. Following coding, we used descriptive statistics to make sense of the larger dataset and to break it down further into subsets, such as content creator categories and platforms. We compared the social media strategies across governments, nonprofits, and influencers and interpreted the associated discourse. Language and imagery were interpreted at creator and platform account levels for information-sharing posts. In other words, we assessed a creator’s characteristics across all platforms and their characteristics for each of their accounts. We looked for continuity in the information shared. While we considered governments and nonprofits as classes, we did not look for consistency across all influencers due to their independent nature.

When all three were divided into groups, the authors started a grounded reading of the memos (Charmaz 2014). Through this reading, we observed certain trends in how influencers and nonprofits interacted with government agencies and vice-versa. Due to the focus on both federal and local government, there were some varieties. For example, Florida’s local governments were nearly absent because the county-based health departments restricted their use of social media to the state’s Department of Health authorized messaging, which was silent on Mpox. Influencers emerged in their absence. Others were overrepresented, such as Portland and Boston. The authors documented how these relationships seemed to be forming among the messiness, creating the framework.

2.3. Ensuring Trustworthiness in Data Collection and Analysis

While recognizing the messy nature of social media, research design trustworthiness was addressed via two techniques. First, the team was trained on the qualitative data analysis tool. Then, we practiced using the codebook on multiple posts as a form of inter-rater reliability checking. This allowed us to corroborate the coding schema and make clarifications where necessary in the codebook (Onwuegbuzie and Leech 2007). For instance, how we defined a set of sequential posts was reconsidered to account for Twitter/X threads, TikTok series parts, and influencer replies to others’ tweets. Second, a graduate assistant supervised the data collection and coding in real time to ensure completeness and accuracy. These steps allowed for the constructivist analysis of the memos generated from the coding (Charmaz 2014).

3. Results

All content creator types posted information-sharing, community-building, and action-oriented content related to the Mpox outbreak. However, they all prioritized information sharing, as indicated in

Table 1. Information-sharing content frequency by content creator group.

	Social Media Posts Sharing Mpox Information
	Primary Focus	Secondary Focus	Total Posts
Government	474	90	563
Nonprofits	218	19	239
Influencers	471	87	558

, as either the primary or secondary focus.4 Twitter/X was mainly used for information sharing, aligning with the platform’s mission at the time, “To give everyone the power to create and share ideas and information instantly, without barriers”. Governments were more active on Twitter/X than influencers or nonprofits. Influencers tended to use Twitter/X to reuse their content from other platforms (e.g., reposting their TikToks or sharing their YouTube videos) or retweet others’ content, but they received more engagement on other platforms. YouTube and blogs were solely used for information sharing and only by one content creator, @MarenHunsberger.

3.1. Themes in Content Creation

During the analysis, we found themes in content topics from each content creator group, as presented in

Table 2. Content topics by content creator group.

Influencers	Government	Nonprofits
Vaccination locations Diagnostic protocols and treatment updates on Mpox Dispelling myths and fighting stigma Personal photos, videos, and descriptions of having Mpox Advocacy	Vaccination locations Diagnostic protocols on identifying Mpox Assurances on viral spread and treatments Dispelling myths What to do if you have Mpox	Vaccination locations News and updates Fighting stigma Advocacy

. Not surprisingly, each group (government, nonprofits, and influencers) had different focuses, though there was some overlap. For example, all three groups discussed vaccination locations. That said, as will be discussed later, when vaccination location information was disorganized, it was an influencer who worked toward providing information on when vaccinations would be available in different parts of the U.S. Updates and news on Mpox were provided by both influencers and nonprofits, mostly based on what the government had been reporting.

Both influencers and nonprofits worked toward fighting stigma while, similarly, the government worked toward assurances on the dangers (or lack of dangers) of Mpox and dispelling myths around Mpox. For example, @thatgaydoctor created a TikTok (also posted on Instagram) explaining that Mpox is not a gay disease but, instead, is spread by close contact, while @doctor_carlton discussed the stigma of Mpox on podcasts.

Lastly, influencers, such as @pozrn, talked about their own experience with Mpox and what it looked like as a way to lessen the stigma of having Mpox. Influencers talked about the experiences of having Mpox by providing pictures and videos showing various ways they experienced Mpox. This personalization of the Mpox experience accentuated the government’s sharing of information on what Mpox is and what it looks like.

One main area of difference is advocacy, in which both nonprofits and influencers participated. This advocacy specifically focused on getting the government to provide more resources, information, or support for people who were experiencing or at risk for Mpox.

3.2. Messy Communication: A Framework

To better understand the ways that information is distributed, we propose a framework to bring order to disorganized communication. Specifically, we present four types of communication realities: vacuum, chaos, control, and flex control. These four types represent how governments, nonprofits, and influencers coproduce information, though the main catalyst is the government, with the influencers and nonprofits following the government’s lead and responding to the government’s actions and inactions.

Messy communication takes into account the complex interactions between governments, nonprofits, and influencers in the way that information is spread on social media. Public administrators battle with oversight and regulations and an ever-changing social media landscape, which makes it difficult to know how to use social media effectively. Through this framework, we can better identify how social media is used and make recommendations for governments and nonprofits to create content based on the communities’ needs. Figure 1 provides an overview of how each of these players interact during the different phases of messy communication.

3.2.1. Communication Approach 1: Vacuum

In a vacuum, the government provides limited or no information. During this time, influencers and nonprofits fill in the void, not always with good information. Influencers may use the vacuum as an opportunity to present their own theories and perspectives, which may or may not be based on reliable sources. Due to being seen as a trusted source, influencers may be seen as a reliable source of information. While conspiracy theories will always be present, in a vacuum, the government is unable or unwilling to correct disinformation from influencers. This vacuum can be present if the government provides information, but it is not targeted appropriately. This could include not reaching out to vulnerable and impacted communities or not reaching out appropriately to said communities.

The impact of a vacuum is that citizens do not receive important information. A vacuum provides an opportunity for misinformation to fester and grow. When this information is provided by the government, they are not seen as a reliable source or provide incomplete information. This may be because the government does not know how to use social media to reach out to minority communities, or they do not prioritize the minority communities’ needs.

A vacuum can exist at any level of government, forcing local nonprofits to fill the gap. In this study, for example, some cities, such as Jacksonville, FL, had limited social media and did not provide information on Mpox for the local gay community. To fill this hole, organizations such as Equity Florida provided support for the gay community. This included a live-streamed town hall event. Nonprofits could not always fill all the holes, and limited information was provided via social media on where vaccinations and treatments were provided.

Nonprofits, such as Fenway Health in Boston, provided their own graphics and infographics for people to understand the realities of Mpox. On 1 July, Fenway Health tweeted out a five-party Twitter/X thread providing information and infographics on what Mpox is, what the vaccination entails, and fighting the stigma of it. During this time, the Boston Public Health Commission provided updates on how many people in the area had been diagnosed with Mpox.

In conclusion, a vacuum existed in how the government dealt with Mpox. Though we can see this absence in some of the information that the federal government provided, the larger vacuum was at the local level. This led to people being unaware of where they could receive vaccinations and treatment. Importantly, this vacuum was sometimes filled by nonprofits and influencers (as well as other levels of government), but not always.

3.2.2. Communication Approach 2: Chaotic

In a chaotic messaging system, the government provides inconsistent or sometimes contradictory information. This can be for a variety of reasons; for example, it might be that information is changing quickly or the government may not be good at explaining why information may be changing. Also, the government may not be proficient at using social media. Either way, the lack of consistent messaging is confusing to vulnerable populations looking for information. In this chaotic system, multiple groups may be providing information, and not all of this information is correct.

In a chaotic system, nonprofits and influencers provide information that may or may not be factual or helpful. They may also share outdated information while trying to be helpful. This information may be connected, in some way, to the government messaging. During this chaotic system, nonprofits and influencers may advocate for having the government provide more focused information. Nonprofits and influencers work with the government to move the chaotic situation into a more organized situation to reach vulnerable and isolated populations.

There were several examples of responding to the chaotic system. For example, Dr. Carlton (@doctor_carlton on Twitter/X) started organizing and tweeting information on vaccination events across the U.S., especially those that were not well advertised. This was due to a lack of centralization around vaccination information (or, as shown in the vacuum, no information at all about vaccinations). He also advocated for states to create centralized spaces to make this information more accessible.

In addition, @thatgaydoctor identified this confusion on his Instagram. In an Instagram post from 4 August, @thatgaydoctor explored the dos and do nots of Mpox and their personal and professional experience with Mpox. By providing this information, they guided individuals during a time when people were not fully sure how to prevent Mpox and what Mpox looks like. This provided a service when the government was not clear on the information provided. Many influencers started talking about their personal experience with Mpox in an attempt to destigmatize the disease. For example, @olliesinha, a TikToker, not only talked about his personal experience but also answered questions. Other influencers, such as @thatgaydoctor on Instagram and TikTok, worked hard to prevent Mpox from being seen as a “gay disease” to cut through the stigma. On TikTok, @MarenMicrobe provided question-and-answer sessions on Mpox when the information was not particularly clear. This provided clarity when information was being put out.

When the information that the government distributed lacked direction for the MSM (men who have sex with men) community, @PositiveRN2020 advocated for better messaging. In his tweet, he talked about how important it is to call Mpox a sexually transmitted infection. This was in response to what he felt was incorrect messaging around Mpox and its transmission. In this chaotic system, @PositiveRN provided information on Mpox that differed from how the government classified the virus.

In a chaotic situation, there is some information coming out from the government. In the case of Mpox, the information from local and federal governments was there but the information was sometimes unclear or disorganized. In response, influencers and nonprofits tried to translate the confusing information and organize information around vaccines to help the gay community understand what Mpox is and how to prevent it. However, nonprofits and influencers lacked coordination, furthering confusion and chaos.

3.2.3. Communication Approach 3: Control

Control would be what you would see in an authoritarian country, such as Russia or Iran, where the government has censorship control over the media (Motamedi 2024; Vendil Pallin 2017). Governments can also control the social media conversation to make certain topics illegal. For example, anti-LGBTQIA+ laws, which are popping up across the world, make it illegal for organizations to even talk about issues relating to the LGBTIQA+ community. In this frame, we will see all three major players (government, nonprofits, and influencers) having the same message. Specifically, nonprofits and influencers will be parroting the government’s message. This is because the government may have control over the nonprofits and censorship over the influencers.

In this study, we found no instances of active control of messaging, unlike the control of COVID-19 messaging in some U.S. states (Blaskey 2021). Though some states and localities decided not to provide information on Mpox, this did not prevent nonprofits and influencers from communicating in the absence of state and local governments. Governmental actors may have ways to censor and limit information they do not want out, but in the ever-increasing social media climate and new ways to get around governments, it may be harder and harder to censor divergent perspectives.

3.2.4. Communication Approach 4: Flex Control

In a flex control system, governments, nonprofits, and influencers work together but are still independent entities. While the government plans a communication strategy, influencers and nonprofits may support it and create their own messaging tailored to their communities. This coordination allows nonprofits and influencers to use government messaging while maintaining autonomy.

The CDC attempted this when they hired Dr. Demetre to run the Mpox response. For example, the CDC tweeted a video on what Mpox is and how to respond if you have a concerning rash. Indeed, from the beginning, by hiring a doctor who is a part of the gay leather community, a subgroup of the LGBTQIA+ community focused on fetish and fetish-wear, the CDC was attempting to create some sort of trust within an isolated community. Working through influencers who are healthcare professionals, the CDC advertised a virtual training on Mpox (tweeted on 25 July) and provided visualizations of what the Mpox rash looks like (e.g., Twitter/X and Instagram, 25 July). These services provided baseline information for individuals and medical practitioners on what Mpox is from a trusted source.

Local governments were also providing flex control through their information system. For example, on 18 August, the Nevada Department of Health and Human Services (NDHHS) advertised a town hall they were co-hosting with The LGBTQ Community Center of Southern Nevada, also called “The Center”. This joint town hall between NDHHS and The Center allowed the nonprofit and government agencies to provide information on Mpox and how the State of Nevada responded to the pandemic. While this is a normal government process, it requires a coordination process (and trust) through social media.

In the flex control, governments, nonprofits, and influencers work together to provide information. This does not preclude nonprofits and influencers from advocating to the government if they are unhappy with the direction. This collaboration allows information to reach vulnerable populations through multiple avenues.

The reality is that government administrators do not always have control over the messaging. We can see this in the discussion of Mpox, where one influencer focused on providing information on where one could get vaccinated when vaccine information was disorganized. Other influencers used social media to destigmatize Mpox and provide education. While the CDC provided information on social media, the outreach was different when we looked at the local communities. In some cities and states, a lot of work was carried out to provide education and information, while others ignored the situation entirely.

3.3. The Benefits and Challenges of Messy Communication

Messy communication, as a framework, helps public administrators understand the complexities of messaging, specifically social media communication. It mimics the reality of the situations that public administrators face. The reality of trying to use social media to get ahead of a situation (or the decision not to use social media) is a tricky one that is constantly changing based on the information available and changing strategies. Two benefits of the messy communication framework are present. First, having multiple avenues of communication can help the government reach vulnerable communities, especially those who might not trust the government. Second, nonprofits and influencers can communicate in a language that matches the community in ways the government is not always able to.

Each of these four approaches will look different during various stages of a crisis. The CDC (Centers for Disease Control and Prevention Mission n.d.) identifies, for communication purposes, the stages of a crisis as pre-crisis, initial, maintenance, resolution, and evaluation. For example, in a pre-crisis stage, a vacuum would mean no treatment or vaccine information was shared, and governments were not preparing for the health crisis. Meanwhile, influencers or nonprofits might share limited (and often incorrect) information. In comparison, having flex control would allow governments to respond with ever-updating information not only to communities but also for influencers and nonprofits to distribute to harder-to-reach communities.

However, challenges are present as well. The issue with this framework is, in many ways, influencers and nonprofits may have a greater impact on minority communities than governments. Influencers and nonprofits may advocate for priorities different from the government’s priorities. These may be professionals, or these may be people who are spreading disinformation. These nonprofits may have a confrontational relationship with the government, advocating for more attention and/or resources on a problem. The lack of governmental control over content creation and distribution creates challenges for public administrators.

4. Discussion

Social media research is an important part of broader public administration scholarship (Feeney and Porumbescu 2021; Zhu et al. 2020; Campbell and Lambright 2020). Our findings support the existing scholarship on the independent behavior of government and nonprofits, as well as Zavattaro and Brainard’s (2019) argument for user-generated content to create more meaningful online encounters. We added to the field by constructing a framework to detangle how social media is used by the various players in public administration, including influencers. Governments, nonprofits, and influencers can work in concert to build public value, or in this case, “linking experts and novices”, to support public health (Zavattaro and Brainard 2019, p. 572). However, some influencers are legitimized experts themselves. Considering the way that influencers can be seen as experts, social media platforms such as TikTok have changed the way information is disseminated, especially to underserved areas (Wang et al. 2023).

Though this study looked at the U.S., we can see this balance internationally. Examples of flex control include Finland, where researchers found that influencers took government communications and made them their own (Pöyry et al. 2022). Furthermore, in Spain, posts around COVID-19 articulated support for official information being disseminated and created social cohesion (El Mundo 2020; Santoveña-Casal et al. 2021). Lastly, in Italy, the government enlisted influencers during COVID-19 to provide information on the importance of masking (Amante 2020). In comparison, control of the media can be seen in countries such as Russia, which even includes blocking access to certain social media sites (Vendil Pallin 2017), and Iran, which blocks all major social media sites (Motamedi 2024).

The reality is that confronting public problems is often messy and unplanned for several reasons. Constantly changing information makes it hard to focus on messaging (Schillinger et al. 2020). A lack of social media savvy staff to reach out to communities or little authority provided to social media staff delays postings (Morgan et al. 2024). A public distrust of government and loud misinformation (and disinformation) can be difficult to challenge (Mergel and Bretschneider 2013; Hilton and O’Leary 2018; Zavattaro and Brainard 2019; Figenschou and Fredheim 2020). Other theories in public administration acknowledge this messiness, including garbage can models (See Zhu and Kindarto 2016) and contingency theories (See Alford 2002; McGrandle 2017). This social media framework allows us to think about the ways that we interact with public organizations. Instead of prioritizing strategy, this framework acknowledges that social media is sometimes messy. In fact, DeMasters et al. (2024) argued that TikTok users prefer the messiness.

Messy communication provides a framework for evaluating and discussing how public administration communication actually occurs. Take, for example, the cacophony on social media when Florida’s Surgeon General released a report on gender-affirming care that misrepresented peer-reviewed medical studies (DeSantis 2022). The report was immediately challenged by the federal government, field experts, LGBTQIA+ advocacy organizations, and transgender influencers. In effort to control the social media space, chaos ensued. The framework operationalizes social media’s multidimensional nature and acknowledges its constantly changing environment. The framework can be used to reconstruct the complex conversation between multiple parties in public administration and their constantly changing strategies. Thus, we build on Zavattaro and Brainard’s (2019) propositions by offering two of our own:

Proposition 1a. 

Influencers will use social media to supplement where they perceive governments as providing insufficient information for their audiences.

Proposition 1b. 

Influencers will leverage their followers’ trust to further government reach.

The messy communication framework is a tool for operationalizing how nonprofits and influencers either work in tandem with or in contrast to the government on social media. This framework pushes forward the question of how these content creator groups can work to provide public information, especially for minority communities or communities that may not trust the government (e.g., the LGBTQIA+ community and immigrant communities). Are the influencers and nonprofits providing their own perspective and information, or are they using government information (or a combination of both)? For example, in our study, Dr. Carlson used government data (the locations of vaccination sites) and organized it in his own way to reach the LGBTQIA+ community. While working with the government data, he informally partnered with the government and acted as an influencer to spread important information to those who may not have seen it.

Governments can incorporate influencers as part of the outreach process. The U.S. Food and Drug Administration (Food and Drug Administration 2022) even has resources on how governments can engage with influencers. These relationships can help governments promote new campaigns or resources and increase credibility and trust. These partnerships could be beneficial for both the government and the influencer.

The CDC’s Crisis and Emergency Communication Plan (Centers for Disease Control and Prevention Mission n.d.) provides specific guidance on how governments should communicate during emergencies. This communication looks different in each crisis stage (e.g., planning, response, and recovery) for local, state, and federal governments as well as nonprofits. Missing from the CDC’s report is working with influencers. Though governments have worked with influencers (see Amante 2020), more guidance on navigating this relationship is needed, especially for local and state governments.

One issue to prepare for is that influencers might be popular with minority communities but may not be so with larger communities. The New York City (NYC) Department of Health and Mental Hygiene (De Blasio Administration Publishes 2017) experienced this rejection when social media users unsupportive of NYC’s LGBTQ Health Care Bill of Rights attacked Dr. Demetre. An online mob accused him of satanism for his “thirst trap,” shirtless Bare it All campaign to encourage patients to be open with their doctors about their sex life and drug use (Daniels 2022, para. 4). We also observed influencers within the LGBTQIA+ community targeting the reduction of risky behavior by incorporating evidence-based crisis response communication strategies. They were action-oriented, timely, and respectful. They demonstrated an effort to increase audience self-efficacy and reduce audience uncertainty (Miller et al. 2021). For example, @pozRN provided important information on how to practice safe sex, get tested, and access vaccines. However, his tagline “Ho responsibly”, while popular within the gay community, may be a turn-off for other audiences.

It is important, when talking about influencers, to adapt the approach to the community, or in this case, to “queer” communities’ expectations (Meyer and Millison 2022). First, make sure that those influencers are credible to the target audience. Second, ensure partnering influencers represent the diversity of the minority communities targeted. Third, prioritize the safety of the at-risk audience over the comfortability of other audiences while following evidence-based crisis response tactics. This framework provides guidance on how researchers can look at the way information is spread through these multiple outlets to create information on important topics.

Proposition 2. 

Social media is a constantly changing force. Governments can enhance their social media capacity by partnering with socially oriented influencers to distribute time-sensitive information.

Feeney and Porumbescu (Feeney and Porumbescu 2021) discussed that government agencies do not have the knowledge and resources to use every social media platform. It is hard to keep up with the ever-changing landscape; as of this writing, Twitter/X has been rebranded as X, and it is unclear if it will still be around within a year (Ocampo 2023) and what, if anything, will take over, with multiple platforms, such as Bluesky, Mastadon, and Threads, vying for people’s attention. However, Twitter/X has been essential in disseminating information to communities (Campbell and Lambright 2020; Moon and Hadley 2014).

Governments may find use in working with nonprofits and influencers to increase the effectiveness of their social media activities. Influencers can serve as network hubs in minoritized communities in ways governments may not. These partnerships can expand the capacity of governments to engage in social media, allowing them to support communities through diverse social media accounts when the government does not have the knowledge or connections. For example, @PozRN documented his personal experience with Mpox in a direct way that broke down barriers and decreased stigma. This framework provides guidance into how governments can evaluate how their relationship with influencers and nonprofits can work and proceed. Governments can look to Finland, Spain, Italy, and the U.S., as they pave the way for this public–private innovation as early adopters.

When preparing practitioners to work in public administration, we hope this study pushes forward the conversation on creating strategies on social media. As things change quickly, public administration can be at the forefront of practitioners understanding the complexities of social media. Instead of being nervous about the messiness, we hope this research encourages practitioners to lean into the disorganization and be aware of it. Openness can help create a communication strategy that is flexible when information changes and helps guide others (i.e., influencers and nonprofits) to share a similar message and respond when misinformation becomes an issue (Mishra and Ashfaq 2023). Through working with nonprofits and influencers, governments are able to reach minoritized communities, as seen in this study, with @doctor_carlton and @PozRN providing information to the gay community. Influencers’ ability to use various social media platforms and knowledge of their changing landscape provide the government with additional avenues to reach diverse communities.

Further research can explore how, in the ever-changing social media environment, government entities work with or against nonprofits, influencers, for-profit companies, and other parties (and vice-versa). For instance, scholars could use this framework to explore how influencers can work with governments and nonprofits for issues that are important to public administration, such as anti-vaping or other public health campaigns. The Australian anti-vaping campaign, which includes eight influencers, provides an excellent source for a case study (Lowrey 2024).

5. Limitations

There are, as with any study, limitations to acknowledge. To start, this study focused on one public health issue based in the U.S., a democratic republic. In every country, the government has different relationships with nonprofits and influencers. Therefore, this framework should be expanded to explore how different countries and regimes moderate conversations on social media. For instance, China’s relationship with TikTok and the effect of that relationship on TikTok in the U.S. has resulted in wide-scale bans of the platform (see Franklin 2023).

Furthermore, with the growth of anti-LGBTQIA+ laws, how nonprofits and influencers interact with the government (or do not) to reach out to the LGBTQIA+ community is important to discuss. Such bans disrupt influencers’ activities. Next, we did not measure the impact of this social media activity on audiences receiving the messages. Lastly, there is no uniform definition for influencers. The definitions vary across disciplines, audiences, and platforms. The influencers included in this study may not align with all definitions of the term.

6. Conclusions

During the summer of 2022, the Mpox outbreak in the U.S. mainly impacted the gay community. Governments, nonprofits, and influencers sent important information to the gay community through social media, including Twitter/X, Instagram, Facebook, TikTok, YouTube, and blogs. Based on our findings and building on previous work on social media, we proposed a structure for analyzing and understanding the way that social media is used in public administration: messy communication (Zavattaro and Brainard 2019; Feeney and Porumbescu 2021; DeMasters et al. 2024; Li et al. 2021). Applying this framework, practitioners and scholars can better examine an ongoing challenge for public administration: social media. We offered two propositions for leveraging governmental and nonprofit partnerships with influencers via flex control. Through these propositions, we aim to guide scholars on operationalizing social media usage. Scholars could use ethnographic methods to observe online behavior and interview content creators (see DeMasters et al. 2024) or use network analysis to assess important vectors and measure reach (see Ocampo 2023). Most importantly, the framework provides guidance to public administrators who rely on social media in their work.