Abstract
This article investigates how social media can enable and constrain civil society organizations’ (CSOs) discharge of accountability. Based on a comparative analysis of the Instagram, Facebook, and Twitter (X) posts of the Swedish Red Cross during 1 year (N = 1014), we propose a framework of affordances that illustrate how platform features, practices, norms, and perceptions about audiences jointly shape the accountability potential of a platform. Accountability is overall more content- than process-oriented, emphasizing visibility of action rather than far-reaching social interactivity. Our study, however, reveals important differences between Instagram and Facebook, on the one hand, and Twitter, on the other. Whereas accountability is more short-term, scripted, and donor-oriented in the former, it is more abstract and ad hoc, with mainly indirect efforts at interactivity, in the latter. Our framework of affordances sheds light on the hitherto under-researched intersection between the literature on CSO accountability and the literature on CSO use of social media.
Introduction
Social media has earned an important role in organizing and communicating in civil society; it has been a vehicle for mobilizing protests (Gerbaudo, 2016) and coordinating volunteers in grassroots initiatives (Kaun & Uldam, 2018) as well as a channel for fundraising (DiLauro et al., 2019), recruiting volunteers (Briones et al., 2011), undertaking advocacy (Guo & Saxton, 2020), and engaging in stakeholder dialogue (Bellucci & Manetti, 2017) in established organizations. Overall, the literature on social media use in civil society illustrates a manifold role of social media, as well as several opportunities and challenges resulting from it.
An aspect of social media in civil society that has earned surprisingly little attention is accountability. Due to their mission complexity, civil society organizations (CSOs) have a long history of engaging with stakeholders and are facing mounting pressures to discharge accountability to their various stakeholder groups (O’Dwyer & Unerman, 2008; van Zyl & Claeyé, 2019). A common idea of accountability is the ability to explain and justify one’s conduct to someone else (Messner, 2009; Roberts & Scapens, 1985) and to allow these others to “ask questions and pass judgment” (Bovens, 2007, p. 450). While various views exist on how accountability may be understood and practiced (Sinclair, 1995), a common interpretation entails the production of formal reports on the use of funds and short-term outputs (Dhanani & Connolly, 2015). Such interpretations of accountability have been criticized for being narrow and accommodating powerful stakeholders such as donors and regulators (O’Dwyer & Unerman, 2008). Alternative approaches to civil society accountability have been discussed, such as providing qualitative accounts or engaging in dialogue with stakeholders (Cuckston, 2022). Meanwhile, social media, with their capacity for rapid and broad (Treem & Leonardi, 2013) dissemination of information to diverse audiences and with the potential for interaction with these (Bellucci & Manetti, 2017), have emerged as important platforms for accountability. Several studies have pointed to the potential for for-profit companies to provide both old and new forms of accounts on social media and to be held accountable by various audiences and to interact with them (e.g. Blankespoor, 2018; Karunakaran et al., 2022; Vestergaard & Uldam, 2022). Considering the increasing expectations on CSOs to discharge accountability, coupled with the potential of social media, the relative absence of studies combining the two areas is surprising.
In the literature on social media in civil society, studies have typically focused on a single platform, such as Twitter (Lovejoy et al., 2012; Nau et al., 2023) or Facebook (Cho et al., 2014; Saxton & Waters, 2014). As prior research suggests (Bossetta, 2018; Matassi & Boczkowski, 2021), however, different social media platforms afford different functionalities or low-level affordances, for example, character limit, possibility to use images and videos, and mode of user interaction. Moreover, studies on individuals’ use of social media underline social norms associated with different platforms (Alhabash & Ma, 2017; Boczkowski et al., 2018; Bucher & Helmond, 2017). These technical and social dimensions could hence be expected to influence how accountability is discharged. However, research on CSOs’ use of social media has not investigated how such differences materialize in practice (Guo & Saxton, 2020).
Therefore, the article aims to undertake a comparative study to explore the role of social media in discharging accountability. The following question guides our investigation:
How can different social media platforms enable and constrain CSOs’ discharge of accountability? To address our research question, we analyze the posts of the Swedish Red Cross (SRC) on their Instagram, Facebook, and Twitter platforms during 2022. SRC is a large professional CSO that pursues several missions and is highly dependent on fundraising. Therefore, steady communication can be expected. The analysis is informed by the idea of affordances. Our study draws on and contributes to the literature on CSO accountability and CSOs’ use of social media.
The rest of the article is structured as follows. First, we present the main tenets of CSO accountability and synthesize findings of prior research regarding how CSOs use social media. Then, we describe the idea of affordances and how we use it. Thereafter, we account for the methods and materials used and we report on the different forms accountability assumes in SRC’s social media channels. Next, we propose and discuss a framework of affordances for understanding civil society accountability on social media. We conclude the article by highlighting our main contributions and suggesting how future research could continue in this comparative vein.
Accountability and affordances
CSO accountability: contents and processes
While CSO accountability may be discussed in various terms, we are primarily interested in the means and processes by which accountability may be discharged rather than in the possible targets of accountability. In reviewing the broader accountability literature, Messner (2009) points to the common distinction between what to include in an account and how to interact with an account. Hence, one central question concerns the content of accounts, that is, what type of information an accountor should provide to an audience to discharge accountability. Another central question concerns the processes by which accounts are produced, that is, how the accountor and the audience negotiate and make sense of the information (Messner, 2009). Critics across contexts have argued that accountability is typically concerned with content and, more specifically, the standardization and even quantification of information (Ebrahim, 2003; O’Dwyer & Unerman, 2008; Roberts, 1991). A concern related to this observation is that a focus on information per se risks creating illusions of transparency (Roberts, 1991), reporting, for example, on measurable outputs at the expense of more complex issues (van Zyl & Claeyé, 2019). This, in turn, could form a distance between the accountor and the audience (Roberts, 1991) and discourage dialogue between parties (Ebrahim, 2003). Researchers have suggested that CSOs should report not only on resources, activities, and short-term outputs but also on more long-term outcomes of activities for beneficiaries, and on the organization’s mission (O’Dwyer & Unerman, 2008). As prior research shows, CSOs tend to account mostly for the short term or success stories (Chu & Luke, 2023; Dhanani & Connolly, 2015). This tendency has been explained by the costs associated with producing information and a fear of not obtaining funds if negative information is presented (Chu & Luke, 2023; Dhanani & Connolly, 2015). Another idea of how to more fully align CSO accounts with accountability can be found in Cuckston (2022), who suggests that accounts need not necessarily be quantified but could build on qualitative descriptions. Cuckston argues that CSOs’ missions do not easily lend themselves to quantified measures and, therefore, demonstrating actions that “make a difference” would suffice in discharging accountability. A few studies have shown that CSOs often mobilize individual experiences when accounting for their activities and outputs in annual reports. Such accounts often center around the organization itself and its employees (Dhanani & Kennedy, 2023). Not only are the organization and its employees portrayed as active and competent, but they are also typically the ones that describe the needs. Beneficiaries are portrayed as vulnerable and passive recipients of the aid. When they are given a voice, beneficiaries often provide testimonials that convey gratitude and serve as receipts of the organization’s output (Dhanani & Kennedy, 2023).
In addition to the content-oriented view of accountability, several scholarly works have explored the more processual side. Roberts (1991) argues that it is in the accountor–audience interaction that accountability emerges, where an informal and close relationship between the parties can be created and focus can shift from the informational content to the demands and expectations of “the other” (i.e. the audience). Similar ideas can be seen in the work by O’Dwyer and Boomsma (2015), who suggest that CSO accountability is about “opening themselves up to scrutiny” (p. 38). Ebrahim (2003) points to the potential of dialogue with both donors and beneficiaries as a means for CSOs to enhance learning and become more inclusive of others’ perspectives in planning activities and assessing performance. However, as Ebrahim (2003) and others (e.g. Madianou et al., 2016) suggest, more interactive modes of accountability would need to address possible power imbalances. The fact that beneficiaries are dependent on the services provided by CSOs and that CSOs are dependent on funds provided by donors and governments is one such source of imbalance that may hamper the fostering of mutually understanding dialogues.
CSOs and social media
Studies demonstrate a wide range of functions that social media may fulfill in CSOs’ external communication, in essence distinguishing between providing information (one-way communication) and interacting with the audience (two- or three-way communication) (Chu & Luke, 2023; Dinh & O’Leary, 2024). While few studies explicitly discuss accountability, some provide insights important for understanding the role of social media in discharging accountability.
In line with a content-oriented approach to accountability, CSOs use social media to report on their program and project activities. This is seen in studies of Twitter (e.g. Lovejoy & Saxton, 2012; Nau et al., 2023) and Facebook (Bellucci & Manetti, 2017). Alexander et al. (2023) distinguish between posts about achievements in terms of fundraising and program activities and posts promoting upcoming activities and projects, and demonstrate a positive correlation between the more backward-looking content and ease of external monitoring by institutional funders. These findings suggest that social media play a role in providing accounts and that such account giving may be perceived by external stakeholders as an important part of accountability.
Relatedly, CSOs use social media to provide information about social, environmental, political, and health issues. Such awareness-raising posts were common in Bellucci and Manetti’s (2017) study, implying that CSOs find it important to inform the audience not only about actual activities and achievements but also about the problems and causes they seek to address with their programs and projects. The formulation of problems has been observed for CSO accountability more generally. Cuckston (2022) describes this as a way of defining the boundaries for what counts as performance in a field where performance is not easily defined and measured. Presenting problems creates a space in which to perform actions that count as performance (Cuckston, 2022). More critical scholars frame such practices as a means of creating legitimacy by identifying needs (e.g. Dhanani & Kennedy, 2023). While we do not view such information as a direct discharge of accountability, we see it as an important framing in which accountability-relevant information may be provided.
Prior research also suggests a role for social media in process accountability. Some (e.g. Chu & Luke, 2023) suggest that posting regularly on social media per se invites interaction with stakeholders, thereby exposing organizations to questions and comments. Others focus on the actual interaction as a means of discharging accountability. Alexander et al. (2023) show that higher levels of user engagement are associated with ease of monitoring by institutional donors. Hence, user engagement—regardless of its nature—may be seen as important. Dinh and O’Leary (2024) show how social media are increasingly becoming venues for beneficiaries of local humanitarian operations to put forth requests and suggestions, and to independently express their vulnerability. However, several studies conclude that CSOs use social media for broadcasting information rather than for soliciting input or entering into dialogue (Bellucci & Manetti, 2017; Lovejoy et al., 2012; Nau et al., 2023; Wang & Yang, 2020). When interaction occurs, it is often limited to superficial and positive comments from stakeholders (Chu & Luke, 2023), and organizations do not necessarily respond to all comments (Bellucci & Manetti, 2017). Chu and Luke (2023) find only a few instances of what they refer to as “meaningful dialogue,” that is, interactions between organizations and stakeholders that include several comments and responses, with more substantive content. However, some would argue that unidirectional posts constitute a basis for building an online community that could engage with more dialogical posts (Lovejoy & Saxton, 2012).
In sum, extant literature indicates that CSOs use social media in several ways that enable the discharge of accountability toward stakeholders. Notably, social media are used for providing information about problems that CSOs seek to address and about programs and projects that CSOs run. Also, social media are used to solicit comments from external stakeholders and to respond to these, although there is some indication that such interaction may be rather limited and superficial. Due to the quantitative nature of many prior studies, however, we know rather little about how the different types of accountabilities are discharged. Furthermore, as extant studies have either focused on a single platform or discussed social media on a more general level, the role of different platforms in discharging accountability is not well understood.
Social media affordances and accountability
Our comparative approach necessitates sensitivity to platform-specific features, norms, and practices. We approach this through the lens of affordances. On a general level, affordances may be understood as “what material artefacts such as media technologies allow people to do” (Bucher & Helmond, 2017, p. 235). More specifically, affordances in social media may be seen as “the perceived, actual or imagined properties of social media, emerging through the relation of the technological, social, and contextual, that enable and constrain specific uses of the platforms” (Ronzhyn et al., 2023, p. 3178). An affordance lens thus draws attention to both material and social aspects of the use of a platform, without assuming a techno-deterministic perspective. Platforms have different material features, manifested in, for example, templates, like buttons and search bars, but also in terms of algorithms that regulate how content is presented to users (Kaun & Uldam, 2018). However, there is no 1:1 correspondence between a material feature and an affordance, that is, between what designers had in mind and how a feature is used (Nagy & Neff, 2015). Affordances are hence relational and contextual in that they are actualized in the interaction between platform features and users (Treem & Leonardi, 2013), where the goals, norms, resources, and preferences of individual and organizational actors jointly shape the understanding of what a platform allows them to do (Ronzhyn et al., 2023). Most importantly, following the seminal discussion by Gibson (1982, p. 411, cit. in Bucher & Helmond, 2017), “affordances do not cause behavior but constrain and control it.”
A common way to hierarchically organize affordances is through distinguishing between high-level and low-level, where high-level affordances are more abstract conditions enabled by technologies, such as “visibility” (e.g. Treem & Leonardi, 2013), whereas low-level affordances are typically associated with more concrete features of specific platforms, such as a like button (Bucher & Helmond, 2017). However, referring to design features as low-level affordances is criticized by Ronzhyn et al. (2023) since it assumes a techno-deterministic perspective and neglects the relational nature of affordances. For this reason, we refer to design features as features, or properties, and not as low-level affordances. We take our point of departure in two affordances that are of particular relevance to our conceptualization of accountability: visibility and social interactivity. Visibility, or similar conceptualizations, has been proposed as a key affordance of social media (Namisango et al., 2022; Ronzhyn et al., 2023). Treem and Leonardi (2013, p. 150) describe visibility as “the ability to make their [organizations’] behaviors, knowledge, preferences, and communication network connections that were once invisible (or at least very hard to see) visible to others” through, for instance, status updates/posts. This is relevant to the idea of content accountability, which rests upon the assumption that accountability may be discharged through rendering visible the activities that CSOs undertake to address problems and causes. Although we acknowledge that visibility as a social media affordance has been formulated in different ways, we believe that this basic definition can be helpful in our analytic endeavor. While some would argue that social interactivity may follow from visibility, it may be seen as a high-level affordance in its own right (e.g. Namisango et al., 2022; O’Riordan et al., 2016). Following O’Riordan et al. (2016) we define social interactivity as “the potential for users to communicate with social connections” (p. 248). Social interactivity is relevant to the idea of process accountability, where interactions between CSOs and their external stakeholders are central. We are thus interested in elements of the use of social media platforms that afford visibility for content as well as elements that afford the possibility to enter dialogue. As shall be seen in the discussion, we use these two high-level affordances to conceptualize in more detail—through a framework of low-level affordances—how the different platforms can enable and constrain the discharge of accountability.
Hence, we use affordances to conceptualize how design features of specific platforms together with user perceptions, norms, and practices create perceived opportunities and constraints for discharging accountability. Importantly, we do not view accountability as an affordance (but see Zheng & Yu, 2016) but as something that emerges through the possibilities and constraints involved in our proposed affordances (cf. Ronzhyn et al., 2023). Studies on CSO accountability have typically not engaged with the idea of affordances (for an exception, see Madianou et al., 2016), yet we believe it will help articulate a theoretical contribution that could advance the literatures on CSO accountability and CSOs’ use of social media.
Methods
To understand how different platforms can enable and constrain CSOs’ discharge of accountability, we opted for a study of a single organization. As the intersection between CSO accountability and CSO use of social media has received limited attention, a single case was deemed valuable as it would allow for exploring novel elements in depth while creating a “bounded system” that reduces the need for additional comparisons (Creswell & Poth, 2016). Moreover, as our affordance perspective assumes that opportunities for action emerge in relation not only to platform features but also to practices, norms, and perceptions of the users on these platforms (Ronzhyn et al., 2023), we wanted to remain attentive to organization-specific characteristics. Our choice of case has several reasons. SRC is one of the largest humanitarian voluntary organizations in Sweden (SRC, 2024), with revenues amounting to 1492 million SEK (~132 million EUR), 100,000 monthly donors, 88,000 members, 22,000 volunteers, and 441 employees in 2022 (SRC, 2023), which is comparable to the other three dominant humanitarian organizations, UNICEF Sweden, UNHCR Sweden, and Save the Children Sweden. As such, the context may be described as accountability-intense. SRC is furthermore well established on social media, with six active accounts and many followers. We therefore expected their social media to provide rich insights into different forms of accountability.
Instagram (~19,300 followers), Facebook (~175,000 followers), and Twitter (~15,500 followers) were chosen since they are the dominant social media of SRC. On its website, SRC links to all three accounts. In addition, SRC links to its LinkedIn account and YouTube channel, yet, the LinkedIn account is mostly directed at recruitment and employer branding, and the YouTube channel differs from other platforms to the extent that systematic comparison is not possible. SRC also has a TikTok account, but it is rarely used and seems to mostly market their secondhand stores. SRC does not maintain a Snapchat account. All posts during 2022 (N = 1014) were collected. For Instagram (rodakorset, N = 189) and Facebook (rodakorset, N = 225), we used CrowdTangle, which downloads page posts and corresponding metadata, including full text, links, information about images and videos, as well as the number of comments, likes, and shares. Posts from Twitter (@RodaKorset, N = 600) were collected through the Vicinitas.io platform, which similarly downloads tweets and corresponding metadata, including full text, links, information about images and videos, as well as the number of favorites, mentions, and retweets, and whether the tweet is an original tweet, a retweet, or a reply. This provided a comprehensive dataset that was still possible to analyze in depth.
To add detail, we manually coded all posts as to whether they included a link to an external site (yes/no) rather than to SRC’s website or websites of Red Cross organizations in other countries. This was relevant for capturing further nuances of content accountability. Also, we manually counted SRC’s share of the total number of comments following on each post (cf. Bellucci & Manetti, 2017) to better understand process accountability. In addition, we manually coded all posts in several more multifaceted dimensions to learn more about other aspects relevant to accountability: the purpose of the post (WHAT), who is speaking (SPEAKER), and geographical area (WHERE). These dimensions and their accompanying codes (described below) were derived in an iterative process where both authors coded a subsample of posts inductively while synthesizing prior studies on CSO accountability and CSOs’ use of social media in parallel (cf. Krippendorff, 2018).
WHAT. This dimension served to illuminate the extent to which social media are used for content accountability. Two codes were particularly relevant here. First, it was important to identify posts that provided information about SRC’s program and project activities (cf. Nau et al., 2023) (coded as Action). Also, posts describing problems and causes without demonstrating SRC’s efforts (cf. Bellucci & Manetti, 2017) (coded as Awareness-raising) were relevant to understand the extent to which SRC can speak of problems without demonstrating action. If posts could not be categorized as Action or Awareness-raising, we coded them as Advocacy (posts advocating for change), Education (posts containing advice to the general public), Merchandise (posts encouraging the purchase of SRC products and services), Cross-promotion (posts highlighting organizations beyond the Red Cross sphere), Information (posts containing “dry” information, such as upcoming events and new appointments), or Photo opportunity (posts with no clear message; usually a group photo). A residual code (N/A) was used for posts that did not suit any of the above, such as holiday greetings and condolences. While not related to content accountability, these categories appear in studies on CSOs’ use of social media and illustrate the wider context of communication where content accountability is discharged. Moreover, all categories are relevant for understanding process accountability, where the focus lies on the interaction between SRC and its audience rather than on whether the post reflects content accountability or not.
SPEAKER. This dimension provided a more fine-grained understanding of how content accountability is discharged, by assigning a voice to each post. In line with research on CSO accountability (e.g. Dhanani & Kennedy, 2023), we identified the occurrence of co-workers and beneficiaries in accounts of SRC’s activities. 1 Our data also demonstrated the simultaneous presence of co-workers and beneficiaries as well as others (e.g. celebrities promoting SRC, external experts, representatives of other organizations). In cases where no specific person was talking, the code SRC was used. The final codes included SRC, Co-workers, Beneficiaries, and Co-workers/Beneficiaries and were applied to all posts, regardless of how the purpose was coded (WHAT).
WHERE. Like the previous dimension, this was aimed at unearthing more details on how content accountability is discharged. Since SRC operates globally, we were interested in how domestic and international concerns are balanced. This dimension was coded as Sweden, Ukraine, or Other international, depending on the geographical area in focus. Ukraine was coded separately since we expected it to occur frequently after the invasion that started in February 2022. A residual category (N/A) was used for posts that were not connected to a geographical site.
Both authors independently coded the entire dataset according to the coding scheme above. An intercoder reliability test demonstrated an average accuracy of around 80%. Our empirical observations are described in the next section, along the lines of content and process accountability. Our reading of the material was informed by the idea of affordances, which served as a “sensitizing device” (Klein & Myers, 1999) that directed our attention to various material and/or social aspects related to the platforms, the nature of SRC’s operations, and emerging practices and norms for communication. Our affordance-based analysis of the material is presented in the section after.
Content and process accountability on Instagram, Facebook, and Twitter
Content accountability: different modes and voices to describe actions
As seen in Table 1, Action posts, that is, posts with information about activities and projects, dominate in all three channels yet are significantly more dominant on Instagram and Facebook than on Twitter. On Twitter, Awareness-raising posts are almost as common, followed by Information and Advocacy posts, which are also rather important categories in this channel. While overall this indicates an important role for social media in discharging content accountability (cf. Bellucci & Manetti, 2017; see also Lovejoy & Saxton, 2012; Nau et al., 2023), it also provides nuance, suggesting more variation and room for speaking about problems that SRC has yet to solve (Awareness-raising) on Twitter.
Purposes of Posts.
How SRC’s actions are described also varies between the channels, depending on the different modes and voices involved. Overall, qualitative descriptions dominate (cf. Cuckston, 2022). Figure 1 illustrates a post depicting the situation in the war-torn Tigray region in Ethiopia. It contains a native video, in which a mother describes that there is nowhere to hide, nothing to eat or drink, and nowhere to sleep. The video then cuts to an ICRC aid worker (pictured in Figure 1) describing the food and medicine being distributed in the region, and then it cuts to images of a truck being unloaded. At the end, the video cuts back to the mother, who says “In all our hardships, the Red Cross has become our hope. There are no words which can describe how much it means.” The accompanying text reads: Rare images from conflict-ridden region Tigray where children and adults lack food, water, and care. Here, we have just delivered food, medicines, and other important help which saves lives and provides hope for [name of the mother] and so many others. We get through, even when it is hard.
While this type of Action post is common, there are posts that include vaguer indications, such as an image of a boy wading in water together with a short description of monsoon rains in Pakistan and the statement “We carry out preventive work in both the short and long run, in order to minimize damage” (Facebook post, 16 September, 2022) (Quotes translated from Swedish by the authors).
“Big help convoy reaches Tigray.” SRC Facebook post, 16 December, 2022.
Video is used in 34% of the Action posts, ranging from a moderate use on Twitter (13%) to a much more frequent use on Instagram (41%) and Facebook (57%). An additional 41% of Action posts—across all channels—contain images. While images may add richness to an account, it is clear that video enables more detailed accounts, making Instagram and Facebook important for showing the activities undertaken by SRC. Video is also important for providing room for individual voices. Nearly a third (31.2%) of the Action posts involve co-workers’ stories, typically describing the situation in a specific venue and the activities and projects undertaken there. Beneficiaries’ voices are considerably less common, but they appear in 6.5% of the Action posts, and they appear with co-workers in 5.3% of the Action posts, typically sharing their experience of the situation and/or the help received (see Table 2).
Voices in Action Posts.
Note. SRC: Swedish Red Cross.
Comparing the platforms, we note that beneficiaries, both alone and with co-workers, are more commonly heard on Instagram and Facebook (Table 3). This seems reasonable, considering that beneficiaries are mainly heard in Action posts, which are much more prevalent in these two channels. Co-workers are more evenly distributed across the channels yet appear on Twitter more often. On Instagram and Facebook, co-workers typically appear in videos recorded at relief effort sites or in written quotes in Action posts. On Twitter, co-workers often appear through retweets of high-level employees’ personal accounts, providing brief textual descriptions rather than vivid multimodal accounts. However, on Twitter, co-workers’ voices are used for several purposes beyond content accountability; of the posts where co-workers speak, Action constitutes only a third (33%).
Voices across Platforms.
Note. SRC: Swedish Red Cross.
In addition to these different modes and voices, we observe the use of links to account for activities and projects. While all channels allow for linking to other pages, this is done on Twitter only, where 15% of Action posts contain links to external pages, for example, media appearances of SRC. This suggests another style of content accountability on Twitter, based on a wider variety of sources.
Not surprisingly, for a humanitarian aid organization, ongoing emergencies have an important place in SRC’s efforts at discharging content accountability. Ukraine is clearly present in all channels, with a notable dominance on Instagram, where Ukraine-related posts outnumber posts about other countries and posts about Sweden, respectively (Table 4). The latter is not surprising, considering the dominance of Action posts—where SRC often demonstrates its efforts in crises—on Instagram (Table 1). Relatedly, we observe that Action constitutes a significant majority of all Ukraine-related posts on Instagram (73%) and Facebook (77%), further underlining the tendency to show action in relation to matters that might be perceived as particularly urgent. On Twitter, Action represents a lower yet significant share (45%) of the Ukraine-related posts.
Geographical Context of Posts.
Process accountability: interaction mainly on Instagram and Facebook and in relation to non-accountability posts
All channels contain interaction in the form of likes, comments, shares, favorites, or retweets—with likes constituting the main part (Table 5). “Interaction” here includes both audience engagement and responses and reactions by SRC. In line with the differing follower counts, interaction levels vary greatly, with the highest levels on Facebook and the lowest levels on Twitter. This suggests a rather varied potential for process accountability between the channels.
Interactions per Type of Post.
Interaction is primarily driven by posts that are not concerned with content accountability. Moreover, Awareness-raising posts—which are indirectly related to content accountability—generate more interaction than Action posts. It is the residual category that generates the most interaction on Facebook and Twitter, which might be the result of a few seasonal greetings in this category. Aside from this broad category, Advocacy generates the most interaction on Instagram and Facebook, followed by Awareness-raising and Action on Instagram and by Awareness-raising and Merch on Facebook. On Twitter, Merch generates the most interaction, followed by Photo ops, where co-workers of SRC appear together with politicians or people from other organizations. Regarding the various modes employed, photos generate more interaction than video (see Table 5). As for interaction related to who is speaking in the posts, no clear patterns can be discerned.
Looking at SRC’s engagement in these interactions, it is quite noticeable on Instagram and Facebook, with 32% and 30% of the total number of visible comments coming from SRC (Table 6a). SRC thus responds to roughly every second comment from the audience in these channels. The comments by SRC are rather similar, typically including a “Thank you” and/or a heart emoji to recognize an individual donation, an explanation of how SRC intends to address the situation in an unfolding emergency, or an explanation of how SRC uses its funds. While Action and Awareness-raising—the two categories relevant to content accountability—are among the most interactive, interaction between SRC and the audience clearly occurs related also to other categories, such as Merchandise and Education. Interaction between SRC and the audience is highest in Ukraine-related posts (Table 6b).
SRC’s share of interactions, related to purpose of post and geographical context.
6b. SRC comments (total), SRC comments as percentage of total comments.
On Twitter, there is very little direct organization-audience interaction. With the caveat that some replies might have been erased post-publication, there are only three tweets in which SRC responds to a reply to one of their own original tweets. As touched upon earlier, SRC mainly uses Twitter for indirect interaction through linking and retweeting.
A framework of affordances for CSO accountability on social media
Returning to our research question regarding how different platforms can enable and constrain CSOs’ discharge of accountability, it is clear that Instagram and Facebook afford a different degree and other forms of accountability than Twitter. These differences will be discussed in terms of a number of low-level affordances, where platform-specific features, norms, values, and emerging praxis interact in creating opportunities for accountability (Table 7). While these affordances are derived from our two high-level affordances, visibility and social interactivity, our comparative approach demonstrates the value of discussing the role of social media in more specific terms.
A Framework of Affordances for CSO Accountability on Social Media.
Content accountability: concrete and well-curated evidence versus abstract and ad hoc telling
Concerning content accountability, Instagram and Facebook afford showability, implying that accounts may render visible rather specific evidence of actions. Video and images are key here as they have the potential to bring the audience close to “what is going on.” Relatedly, Instagram and Facebook afford individualizability, as video and images enable the visibility of specific persons as evidence of action. In line with prior research (Dhanani & Kennedy, 2023), beneficiaries are often used to provide testimonies to the actions taken by SRC. However, the dominance of co-workers, and also SRC as an organization-level voice, suggests that evidence of actions is largely provided from SRC’s perspective. There may be several reasons for this, for example, related to the convenience of using the voices available in the organization, particularly during the pandemic, and related to critiques about depicting beneficiaries as passive recipients of the CSOs’ services (e.g. Dhanani & Kennedy, 2023). Finally, Instagram and Facebook afford urgeability, meaning that accounts may render visible actions in situations where an immediate response is considered particularly important. These affordances are clearly related to perceptions of the audience. Not only are there many followers on Instagram and Facebook, they are clearly expected to be “common people” compared to the more elite audience on Twitter (Hedman, 2020). Appeals for donations and promotions for merchandise are much more frequent in these channels, indicating a desire to speak to individuals willing to make a difference. Rendering SRC’s actions visible in areas that are currently in the spotlight is likely important also for showing that SRC acts in line with one of their strategic goals—“we rapidly reach out to people in immediate crisis” (SRC, 2023, p. 4). The absence of such evidence might evoke questions about what “SRC is actually doing” (a type of comment sometimes observed in SRC’s social media). While Ukraine dominates the accounts of SRC’s actions, the provision of evidence of action in emergent situations is more generally observed in the two channels. Content accountability is thus rather short-term oriented, something that is probably reinforced by emerging norms on these platforms. For example, Instagram is known to be characterized by a sense of immediacy (Duffy et al., 2017). Hence, there is a risk that the challenges of accounting for more long-term outcomes (cf. Dhanani & Connolly, 2015; O’Dwyer & Unerman, 2008) may be perpetuated as accountability is discharged through social media. In essence, presumptive donors are assumed to want evidence of how their money is used. Research on institutional donors (Alexander et al., 2023) suggests that frequently posting information about activities and projects on social media serves to ease donors’ demands for accountability. Similar ideas seem to be at play here, yet in relation to individual donors. This will be elaborated later in this section.
In contrast, Twitter affords tellability, implying that accounts may make actions visible through short texts describing the current state. The character limit (in 2022, 280 characters) is probably important for this affordance. However, the character limit could be circumvented by using screenshots (which is sometimes observed in our data), video, and images, for example. Our data reveal a limited use of video and images on Twitter compared to Instagram and Facebook, which suggests that there are considerations involved beyond platform features. The possibility of circumventing the character limit relates to the second affordance that we propose for Twitter: assembleability. Assembleability implies that accounts may render actions visible through the seemingly improvised pulling together of sources of evidence. Contrary to the seemingly curated combination of text, video, and images on Instagram and Facebook, assembleability thus involves an ad hoc dimension where external links to media appearances and retweets from co-workers’ private accounts and external actors constitute disparate pieces of evidence of actions. This way of discharging content accountability probably reflects ideas of an audience that is not primarily interested in what SRC is really doing in response to immediate crises (cf. Hedman, 2020). The fact that Action posts are relatively less prevalent on Twitter and on an equal footing with Awareness-raising posts also indicates that SRC expects fewer demands related to content accountability and, possibly, also to process accountability. The latter will be elaborated below. In essence, Twitter enables SRC to speak about problems without immediately presenting solutions, which is probably more due to the perceived target group than to the materiality of the platform.
Process accountability: scripted responding versus indirect interactions
As for process accountability, Instagram and Facebook afford diverse engageability, meaning that audience engagement may unfold in relation to several types of posts. The fact that Action posts are not the main driver of engagement may indicate that these fulfill an information need in the audience and therefore do not evoke as much response as posts expressing an opinion (Advocacy), pointing to a problem that needs to be solved (Awareness-raising), or encouraging purchasing (Merch). These categories may be perceived as more open-ended than the rather information-dense Action posts that show that things are “really” happening. As discussed earlier, frequent posts about activities and projects might soften donor demands for accountability (Alexander et al., 2023). The fact that video is not a major driver of audience engagement could possibly be discussed in similar terms, that is, that they provide details that may fulfill an information need. An alternative interpretation could be that video demands more time of the audience and therefore is not engaged with to the same extent as images—an interpretation that is interesting considering the prevalence of video in SRC’s feed. This may provide some explanation for why video and—consequently—Action posts are not even the main drivers of likes. Regardless of interpretation, our study illustrates the value of investigating actual engagement rather than interactional intent (Lovejoy & Saxton, 2012; Nau et al., 2023). While Bellucci and Manetti (2017) found that posts that involved amusement or demands for feedback generated more comments, our observations are more ambiguous, both in terms of what types of posts encourage audience engagement and whether posts can be understood as either having interactional intent or not. It seems that opening oneself up to scrutiny (O’Dwyer & Boomsma, 2015), that is, approaching a more process-oriented form of accountability, may be enabled by posts that are not accounts of actions nor intended to entice engagement. This is interesting in light of the increasingly pervasive ideal of content accountability in civil society (Cordery et al., 2019; O’Dwyer & Unerman, 2008). Whether this audience engagement contributes to the enactment of process accountability will be discussed below.
Furthermore, Instagram and Facebook afford scripted respondability, implying that the organization may engage with audience comments in a standardized manner. As “Sweden’s largest humanitarian voluntary organization” (SRC, 2024), they are likely to feel an obligation to respond to comments. While Bellucci and Manetti (2017) found that organizations more often responded to comments that appeared in relation to dialogue-oriented posts, we observe organization-audience interactions across a range of posts. The fact that Ukraine-related posts generate relatively extensive interaction between SRC and its audience could be another indication of the perceived importance of appearing accountable concerning particularly urgent matters. Whether organization–audience interactions can be seen as a way of discharging accountability—specifically, process accountability—is unclear, however. Many audience comments do not reflect an attempt at discussing, challenging, or negotiating meaning with SRC (cf., Ebrahim, 2003; O’Dwyer & Unerman, 2008) but rather serve to inform SRC that a donation has been made or to validate the importance of SRC’s work. This observation is in line with prior studies (Bellucci & Manetti, 2017; Chu & Luke, 2023) suggesting an overall positive tenor of audience comments. Inevitably, such comments lead to little more than brief standard responses by SRC. The strong focus on content accountability may render process accountability more positive and superficial, failing to live up to ideals of being open to scrutiny (O’Dwyer & Boomsma, 2015) or being more inclusive of others’ perspectives (Ebrahim, 2003). Whether this is the explicit intent of SRC or not is beyond the scope of this article. Even for more challenging comments, related to priorities and the efficient use of funds, SRC responds in a rather standardized manner, not seldom referring to more detailed information on their webpage. Standardization may hence apply not only to content accountability (Ebrahim, 2003) but also to process accountability. While such interactions should not be interpreted as a lack of accountability, they underline the difficulty associated with developing “meaningful dialogue” on social media (Chu & Luke, 2023), suggesting that ideals described in the accountability literature and in communication research may not be easily transferred across contexts. Earlier studies (e.g. Lovejoy & Saxton, 2012) suggest that purely information-oriented posts could serve to build a community with which the organization could later interact. We believe that this may still be the case, yet our study elaborates on this view by highlighting the possibility that content accountability may condition and ultimately outweigh the importance of process accountability.
In comparison, Twitter affords indirect interactability, meaning that the organization may engage with others (e.g. politicians or other organizations) through retweets, links, and hashtags. This is interesting, as all three channels offer similar features. Still, Twitter affords limited engageability and respondability compared to Instagram and Facebook and, conversely, Instagram and Facebook afford little indirect interactability. This may have to do with the significantly lower numbers of followers on Twitter but may also reflect the very tendency of less interaction observed on the platform compared to, for example, Facebook (Manetti et al., 2017). Twitter’s character limit is likely important for this affordance (cf. Lovejoy et al., 2012). However, as discussed earlier, SRC seems to hold different perceptions about the audience on Twitter in regard to their information needs and their reasons for being present on the platform. Extending this line of reasoning to interaction suggests that direct dialogue in the form of comments and responses to these is perceived as less important as they may not fulfill a direct information need. The relatively low level of interaction could also explain why Twitter enables SRC to speak about problems without presenting the actions taken to address them; they do not expect accountability interactions to unfold to the same extent as on Instagram and Facebook.
Concluding remarks
Our study has provided several insights into how social media can enable and constrain CSOs’ discharge of accountability. Distinguishing between content and process accountability and, furthermore, drawing on the idea of affordances, we have proposed a number of low-level affordances that illustrate the accountability potential of different platforms. Overall, content accountability prevails, meaning that social media, first and foremost, afford CSOs to render their actions visible in a range of ways. While social media also afford more interactive approaches, that is, process accountability, this is enacted to a lesser extent. Our platform comparison reveals interesting differences between Instagram and Facebook, on the one hand, and Twitter, on the other hand, suggesting that platform features, perceptions about audiences, and emerging practices and norms jointly shape the possibilities and limits for how accountability is discharged. Despite seemingly ample possibilities for both content and process accountability across channels, through high-level affordances related to visibility and social interactivity, our study demonstrates a tendency toward short-term and donor-oriented accountability with scripted interactions on Instagram and Facebook and more abstract, ad hoc, and problem-oriented accountability with only indirect interactions on Twitter.
While previous studies have indicated a role for both Facebook (Alexander et al., 2023; Bellucci & Manetti, 2017; Chu & Luke, 2023) and Twitter (Lovejoy & Saxton, 2012; Nau et al., 2023) in discharging content accountability, our more qualitative and comparative study provides more details and nuances and so adds to current knowledge of the possibilities and constraints of social media. Likewise, our study extends the findings of prior research as relevant to process accountability (e.g. Bellucci & Manetti, 2017; Chu & Luke, 2023; Nau et al., 2023) by suggesting that organization–audience interactions may be scripted or indirect depending on the platform and, furthermore, conditioned by the prevalence of content accountability. By articulating and discussing the affordances involved in CSOs’ discharge of accountability on social media, we address the hitherto under-researched intersection between the literature on CSO accountability and the literature on CSOs’ use of social media.
Aware of the limits of single case studies, we encourage additional platform comparisons in other CSOs. While fundraising organizations constitute an accountability-intense context and therefore warrant further study, we suggest greater diversity to enable more nuances to emerge. Future research could focus on fundraising organizations with a longer time horizon, such as those funding research on diseases, to learn about the relevance of the low-level affordances identified in this study. Considering our focus on social media data, we would also encourage the inclusion of experiences and considerations of CSO co-workers, collected through interviews, surveys, and analysis of communication policies. This would make an important contribution to our understanding of affordances related to the use of different platforms, and also in regard to who are the targets of accountability. Finally, to better understand the conditions for process accountability, future research could more closely investigate the nature of accountability in organization–audience interactions on social media and also extend this perspective by taking into account the fragmented and multi-platform ways in which audiences engage with humanitarian communication (cf. Kyriakidou, 2021).
Footnotes
Acknowledgements
The authors would like to thank two anonymous reviewers for valuable insights. The text has undergone language-editing by a professional firm.
Author contributions
Both authors contributed equally to the article.
Consent to participate
Not applicable (see below).
Consent for publication
Permission was obtained from the Swedish Red Cross and from the person depicted in the image to use the image presented in the article.
Ethical consideration
The study is based on secondary data that is publicly available. The primary level of analysis is the communication of an organization in its official social media channels, what purposes it is used for, which modes of communication (e.g. images, videos, and links) are employed, which geographical areas it concerns, and which groups (e.g. employees, beneficiaries, and others) that are given voice in the communication. Engagement by private individuals in relation to this communication, while also being publicly available in these channels, is analyzed and expressed on an aggregated level in terms of number of likes, comments, and shares following a social media post. Ethical approval and informed consent are not required. Care was taken to ensure no personal data were used in the article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by Handelsbankens forskningsstiftelser (grant number P19-0021).
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.