Stakeholder Engagement and Public Information Through Social Media: A Study of Canadian and American Public Transportation Agencies

Abstract

This study uses theories on dialogic accounting to assess whether online interaction through social media is used as a mechanism of public information and stakeholder engagement by Canadian and American public transportation agencies. We embraced a quantitative methodology in which content analysis was performed on the Facebook and Twitter accounts of 35 transit operators in Canada and the United States. We categorized the contents of 1,222 Facebook posts and 2,615 tweets, assessed which level and what type of interaction was effectively reached for every category, tracked whether and how agencies reply to comments on their posts, and assessed the general tenor of the discussion. Our results show that public transportation agencies often take advantage of their presence on social media to provide the public with information on their services and to perform activities associated with stakeholder engagement. However, we have found some significant differences in the utilization of social media by public transportation agencies, all of which are discussed in the “Conclusion” section of this article. Twitter is most often used for public information messages, while Facebook appears to be used more to publish content in a dialogic perspective that creates two-way, collaborative conversations with users. In terms of practical implications, our study suggests that a broader and more continuous commitment to interaction between users and stakeholders on social media would create new opportunities for improving transparency and, indirectly, the services of public agencies.

Keywords

stakeholder engagement social media dialogic accounting public management public transportation

Introduction

Social media applications (e.g., Facebook, Twitter, YouTube, Flickr, etc.) are enabling citizens and private/public organizations to openly and freely interact on the Web (Ma, 2014). According to the extant literature, by 2005 and 2006, strategies related to the Internet had shifted from a primary focus on information to a focus on communication and cooperation (Fuchs, 2008). Some scholars like to designate this transformation as the emergence of “Internet 2.0” or “Web 2.0.” The advent of Web 2.0 not only reorganized the ways in which private and public organizations collected information, but it also redefined stakeholders’ expectations. Social media, for instance, often encourage innovation and improved transparency (Bonsón & Ratkai, 2013; Meijer & Thaens, 2010). Moreover, stakeholders can act as partners and co-creators in these virtual spaces, rather than simply clients or users (Chua, Goh, & Ang, 2012; Manetti & Bellucci, 2016).

These new tools of dialogic communication have opened up new possibilities for public agencies to connect with their stakeholders by allowing them to receive real-time feedback on organizational announcements and engage in conversations. Although one-way communication is still the most common form of messaging strategy adopted by organizations on social media (Brainard & Edlins, 2015; Waters & Jamal, 2011; Xifra & Grau, 2010), attempts to encourage interaction between organizations and users are becoming increasingly popular (Meijer & Torenvlied, 2014; Rybako & Seltzer, 2010). A diverse array of actors in the public sector have adopted and invested in these technologies to guarantee timely information, improve public service quality, and encourage greater participation in various societal debates. Furthermore, social media are expected to enhance participation, learning, and knowledge production in government settings, challenging traditional boundaries of authority (Feeney & Welch, 2016). However, in some cases, the literature shows that although citizens are responsive, interaction levels as a whole remain low due to the non-responsive behavior of public agencies. As a result, it is easy to conclude that although the very existence of agency–stakeholder dialogue is promising, very little of it is collaborative in nature (see, for example, Brainard & Edlins, 2015).

Many transit agencies have begun to incorporate social media into their planning, marketing, and communication strategies. Social media use is valued because it allows public transportation officials to communicate with riders, reach out to potential riders, develop stronger connections with the local community and NGOs, recruit new employees, and improve the agency’s image (Transit Cooperative Research Program [TCRP], 2012). Some transit agencies also use social media applications to engage in customer service activities and obtain feedback from stakeholders on services and programs.

Social media can also pose specific challenges for transit agencies, including content management and strategies for addressing online criticism, estimating resource requirements (particularly staffing) for managing these applications, and developing techniques to measure their costs and benefits. Agencies also have to manage legal and security concerns, including online security, privacy protections, and complying with requirements for transparency and records retention.

In 2012, the TCRP explored the use of social media among American and Canadian transit agencies. The TCRP study highlights some aspects of social media use that require additional research. For instance, although industry experts believe that having a social media policy is critical, a majority of transit agencies have no such policy in place. According to TCRP (2012), some of the main motivators that help explain the adoption of social media by transit agencies include the following:

Public information—providing the public with information about services, fares, long-range planning projects, and all timely updates to share real-time service information and advisories with their riders;

Potential stakeholder and public engagement—taking advantage of the interactive aspects of social media to connect with their customers, employees (e.g., recognizing current workers and recruiting new employees), and local communities in an informal way;

Entertainment and other ancillary functions—entertaining their riders through songs, videos, and contests, and establishing a rapport with users and stakeholders.

Furthermore, most of the agencies in the TCRP study measured the effectiveness of their social media activities by using built-in metrics, such as counting “friends” or followers and using third-party applications such as Google Analytics (TCRP, 2012).

In this study, we analyze the Facebook and Twitter pages of the same 35 transit agencies that were examined in the TCRP study. This is done to understand the level of interaction—especially stakeholder and public engagement and public information—of Canadian and American transit agencies. Unfortunately, scholars have not focused much energy toward understanding why public transportation agencies adopt social networking tools (Li & Feeney, 2014). We will fill this gap in the public management literature by using theories on dialogic accounting, many of which have been embraced by scholars in recent years (see, for example, Bebbington, Brown, & Frame, 2007; Brown, 2009; Brown & Dillard, 2013a, 2013b; Dillard & Ruchala, 2005; Manetti & Bellucci, 2016).

Our exploratory research question involves determining whether or not online interaction through social media represents a mechanism of public information and stakeholder engagement that encourages citizen participation (Bebbington, Brown, Frame, & Thomson, 2007). We are especially interested in how public transportation agencies in the United States and Canada use social media accounts to encourage citizen engagement and spread public information. We ask whether public transportation agencies use social media to distribute information or to try to interact with their stakeholders, and whether or not one can detect forms of dialogue that might make meaningful collaboration with stakeholders possible.

To explore these questions, we will

Investigate the role of social media in creating a model of authentic dialogic accounting and interaction;

Conduct a content analysis of the Facebook and Twitter pages of several transit agencies in the United States and Canada, with specific reference to the functions of public information and stakeholder engagement;

Reach a conclusion on the contribution of social media in the context of American and Canadian public transportation agencies, highlighting the limits of the present research and avenues for future research.

Literature Review

Some of the literature suggests that public administrations should assume at least partial responsibility for citizen engagement to obtain a more responsible form of government (Box, 1998; J. V. Denhardt & Denhardt, 2011; King & Zanetti, 2005). Indeed, many scholars have examined the extent to which public administrations engage citizens (Nabatchi, 2010; Roberts, 2008), paying special attention to how these interactions facilitate public learning and understanding (Ramanadhan, Mendez, Rao, & Viswanath, 2013; Ventriss, 1989/2008); reinforce a sense of community (Etzioni, 1993; Nalbandian, 1999/2008); improve responsiveness (Rosener, 1978/2008); empower citizens (J. V. Denhardt & Denhardt, 2007, R. B. Denhardt & Denhardt, 2000/2008); provide trust in public administration and reduce skepticism and apathy among the citizenry (King, Feltey, & Susel, 1998/2008).

Nabatchi (2010) argues that public administrations have the responsibility to address the citizenship and democratic deficits and the duty to engage citizens within the structures and patterns associated with new public governance (Osborne, 2010). This latter concept, in particular, supports the idea of a pluralist state in which multiple factors inform the policy-making process and many inter-dependent actors agree to deliver public services (Haveri, 2006; Osborne, 2010). The design and evaluation of inter-organizational relationships are crucial to governing the entire process, especially when it comes to building trust and relational capital (Bovaird, 2006; Greve, 2015; Teicher, Alam, & Van Gramberg, 2006).

Social media have become one of the most important instruments of public engagement, inter-organizational relationships, and public information in recent years (see, for example, Kent, Taylor, & White, 2003; Park & Reber, 2008; Porter, 2001; Rybako & Seltzer, 2010; Unerman & Bennett, 2004). Online social media can be defined as “a group of Internet based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content” (Kaplan & Haenlein, 2010, p. 61; see also Cormode & Krishanmurthy, 2008). It is basically an umbrella term describing different types of applications, such as collaborative projects (e.g., Wikipedia), blogs/micro-blogs (e.g., Twitter), content communities (e.g., YouTube), social networking sites (e.g., Facebook), virtual game worlds (e.g., World of Warcraft), and virtual social worlds (e.g., Second Life; Kaplan & Haenlein, 2009, 2012).

Meijer and Thaens (2013) show that the adoption of social media in public administrations (particularly governments) is often inspired by four strategies:

A “push strategy” in which social media are used to broadcast existing web content through other social media channels (Mergel, 2012). At this stage, there is no interaction and social media are used to communicate information.

A “pull strategy” in which the organization uses social media to attract users to the website and provide them with new information. At this stage, public agencies try to pull information from citizens through social media channels, but the interaction level is pretty low because the organization rarely replies to user comments (Mergel, 2012).

A “networking strategy” in which “the use of social media tools is highly interactive with a lot of back and forward between the agency and its diverse constituencies” (Mergel, 2010, p. 10). Public agencies that adopt this strategy often participate in networking and interactive engagement in both an active and passive manner.

A “transaction strategy” (which is rarely observed, according to Mergel, 2012, p. 283) in which public services are actually provided through social media applications.

As Points 2 through 4 suggest, social media applications appear to be particularly well-suited for public information and potential stakeholder engagement, as they make it possible to interact with a large group of people, especially users, employees, NGOs, and local communities. According to Bekkers (2013), social media give users the opportunity to share ideas, opinions, pictures, videos, web links, and other content. Social media can also allow organizations and individuals to exploit people, information, and other sources of data that are present on the Internet by bringing them together to address specific goals or problems. The challenge here is to bring these diverse perspectives together so that new ideas and perspectives can emerge (Surowiecki, 2004), a task that has been made easier in recent years due to the emergence of smartphones, tablets, and notebooks (Bekkers, 2013).

Of course, social media do not always serve as a neutral communication tool for public service providers because goals and actual results do not always converge (Bekkers & Homburg, 2005). This is due to the fact that public service providers often have aims that are rather specific and context-driven, resulting in a host of unintended consequences (Dawes, 2008). Furthermore, the manner in which public agencies manage their social media accounts can often facilitate, slow down, and even prevent web interaction.

However, some of the scholarly literature confirms that public agencies are often motivated to use social media because they believe it represents a sort of institutional duty, because their “competitors” and other public actors have one or more active accounts. These types of arguments suggest that social media use is tied up with ideas on legitimization (see, for example, Brainard & Edlins, 2015). Thus, although it is promising that some public agencies interact with various stakeholders, sustaining multiple-response interactions is often difficult because the organization itself simply ceases to respond.

This study explores the utilization of social media—most notably, Facebook and Twitter—as potential means of dialogic conversation among various public transportation agencies and their stakeholders (Lovejoy, Waters, & Saxton, 2012; Swift, Owen, & Humphrey, 2001). Emphasis is also placed on the opinions and expectations of stakeholders (even when they diverge from the organization’s point of view) and on how public transportation agencies reply to the posts and comments of users. We assess the extent to which public transportation agencies use Facebook and Twitter accounts to create debate and interaction (Unerman & Bennett, 2004), touching on a wide array of topics, including grant-making policies, program funding, and social responsibility issues. We also try to determine how social media activities can help these organizations better define their short- and long-term strategies and spread information among the citizenry.

In the public transportation sector, crowd sourcing plays an important role in generating ideas on transit planning (Brabham, 2008; Kraemer & King, 1986; Pfeffer & Salancik, 1978). In other words, social media usage has the potential to become an effective tool for reaching out to a large number of stakeholders, thus creating dialogue between stakeholders and the organization in such a way as to complement monologic (one-way conversation), dialogic (two-way conversation), or polylogic (among the stakeholders themselves) accounting systems (Bebbington, Brown, & Frame, 2007; Brown, 2009, Brown & Dillard, 2013a, 2013b; Dillard & Ruchala, 2005).

Stakeholder engagement plays a fundamental role in every dialogic accounting process because it allows for a more pluralist expression of public interest, thus “mitigating the dominance of instrumental rationality” (Dillard & Ruchala, 2005, p. 621). According to Brown (2009) and Brown and Dillard (2013a, 2013b), recognizing a diverse array of ideological orientations, enabling access for non-experts, ensuring effective participation, and being attentive to power relations are core principles of dialogic accounting.

Monologic accounting, by contrast, operates under the assumption that the informational needs of investors can affect the values and principles of accounting and reporting systems. As Brown (2009) explains, “monologic accounting also reflects a finality orientation; the ‘facts speaking for themselves” (p. 316). Recent scholarly literature has tried to foster accounting practices that are more receptive to the needs of a “multi-voiced” plural society (Brown & Dillard, 2013a, 2013b), taking into account stakeholder values and interests (Brown, 2009).

These practices are often referred to under the umbrella term dialogic accounting, a concept that recognizes various points of views and refuses to regard capital markets and investors as “priority” stakeholders. Dialogic accounting, in short, rejects the idea of a universal narrative, preferring to think of institutions as being beholden to diverse perspectives and the interests of a wide variety of stakeholders.

Although social media are not normally recognized as an accounting tool, it has the potential to support dialogic accounting systems by providing valuable information on what stakeholders expect from each organization in terms of quantitative or qualitative (narrative) information processing. Indeed, there are some aspects of social media that are capable of altering the relationship between both politicians and public managers and public managers and citizens (Greve, 2015; Landsbergen & Park 2011; Mergel, 2012, 2013). The use of social media in the public transportation sector could have interesting implications in terms of citizen/user participation, distributing information about transit problems, fares, and long-term projects, employee recognition, and even entertainment (TCRP, 2012).

Another possible issue linked to the use of social media in public agencies involves “public-making” (Newman & Clarke, 2009), a process that seeks to address the divergence of opinion among different publics and “new groups” of citizens. Because different groups of people require different services, public transportation agencies often face huge challenges in meeting the needs of a diverse citizenry (Farazmand, 2012).

However, social media’s value in terms of conducting stakeholder engagement often depends on the concept of stakeholder engagement itself. It is necessary to point out, in fact, that stakeholder engagement and stakeholder management are two distinct processes (M. B. E. Clarkson, 1995; Svendsen, 1998; Waddock, 2002). On one hand, stakeholder management foresees to manage stakeholders’ expectations and the claims they support in accordance with their salience (Mitchell, Agle, & Wood, 1997), also balancing these various positions (O’Dwyer, 2005); on the other hand, the stakeholder engagement implies that organizations try to engage primary stakeholders in decision-making processes, making them participants in organization management, sharing information, dialoguing, and creating a model of mutual responsibility. As a result, the main feature of stakeholder engagement is not to encourage the mere involvement of stakeholders to “mitigate” or manage their expectations, but to create a network of mutual responsibility (Andriof, Waddock, Husted, & Rahman, 2002, p. 15; Manetti, 2011; Unerman & Bennett, 2004; Voss, Voss, & Moorman, 2005; Windsor, 2002, p. 138) in which stakeholders are able to interact with each other.

Stakeholder and citizen engagement is conducted to attain two forms of interaction or dialogic communication:

A deliberative, general consensus (Laughlin, 1987, 2007; Power & Laughlin, 1996) based on Habermas’ “ideal speech situation”—a communication among stakeholders in undistorted conditions (Habermas, 1984, 1987, 1989) that can be built in a “public sphere,” “a discursive arena that is home to citizen debate, deliberation, agreement and action” (Villa, 1992, p. 712; Dahlberg, 2005)—on what information and data should be disclosed in the report. When applied to the corporate arena, the result of “an open, honest and unbiased ideal speech situation debate among all stakeholders should therefore lead to the acceptance by all stakeholders of a democratically determined consensus view of corporate responsibilities” (Unerman & Bennett, 2004, p. 691). The deliberative approach can result in both negative and positive outcomes because consensus can never be guaranteed in the “public arena.”

A collection of divergent socio-political views in an agonistic perspective, highlighting the unavoidable values and assumptions associated with different accounts and recognizing the need for multiple engagements between different actors across various political spaces (Brown & Dillard, 2013a, 2013b; Gray, 2002; O’Dwyer, 2005). This perspective recognizes the need for multiple engagements between different actors across various political spaces (Gray, 2002; O’Dwyer, 2005) based on an agonistic model of participation (Brown, 2009; Brown & Dillard, 2013a, 2013b; Dillard & Brown, 2012; Dillard & Roslender, 2011;).

In the deliberative approach, stakeholder engagement is necessary for defining the general consensus among diverse stakeholders. Supporters of the agonistic approach, meanwhile, suggest that stakeholder engagement helps synthesize the different points of views found among groups that have a diversity of interests.

The quality of stakeholder engagement under either a deliberative or an agonistic perspective is affected by several factors (Dahlberg, 2001):

Discourse is often driven by government and corporate interests and, consequently, is not completely independent from state and economic power.

Reflexivity, which is understood as a process of critically reflecting on and changing one’s position when faced by better arguments, is often a marginal part of online deliberations (Streck, 1998) because few online users listen respectfully and come to an understanding of their interlocutor’s positions.

Many social media discussions are characterized by intolerance and a lack of respect between users.

User anonymity on the web can seriously affect the reliability of online interactions.

Discourse tends to be quantitatively and qualitatively dominated by certain users and groups (e.g., excessive posters and users who enjoy higher social status).

Social media can also be used as powerful instruments of legitimization (Bonsón & Ratkai, 2013) rather than a means of creating authentic dialogue and cooperation. Legitimacy theory suggests that a social contract exists between individual organizations and society (Deegan, 2002, 2006; Suchman, 1995). This means that an organization can conduct its activities in a manner that is both socially acceptable and does not necessarily follow stakeholders’ expectations. Thus, organizations can voluntarily report and communicate over social media according to the expectations of society (Pfeffer & Salancik, 1978). Scholars who buy into the legitimacy perspective suggest that organizations use these instruments of external accountability to influence (or even manipulate) stakeholder perceptions (Coupland, 2007; Deegan, 2002; Patten & Guidry, 2010), to reduce their external costs, and diminish pressures being imposed by society or regulators (Adams, 2002; Ballou, Heitger, & Landes, 2006; Caron & Turcotte, 2009; Tate, Ellran, & Kirchoff, 2010). Voluntary information is disclosed for strategic reasons rather than on the basis of any perceived responsibilities. Voluntary disclosure through social media can thus enhance an organization’s legitimacy, elevating its image and perception among various members of society and external stakeholders, especially when using external accountability systems (P. M. Clarkson, Overell, & Chapple, 2011). This has led scholars to question whether social media are used by organizations for legitimizing their presence within society and changing their reputation among stakeholders or rather for creating a system of dialogic—although not necessarily convergent—debate on public issues.

This study adds to the literature on public services by determining whether social media—especially Facebook and Twitter—act as reliable tools of public information and potential citizen engagement (be it in a deliberative or in an agonistic perspective) or whether they are just another means of attaining legitimization. We believe that the literature is lacking in terms of the role social media plays among public services, thereby revealing a gap in knowledge that hinders the development of such studies and their empirical applications.

Methodology

To answer our exploratory research question, we opted for a quantitative methodology based on content analysis, a research technique based on the objective, systematic, and quantitative description of the manifest content of communication (Berelson, 1952). Content analysis is a flexible approach to the examination of various media, documents, and texts, one that seeks to quantify content in terms of predetermined categories and in a systemic and replicable manner (Bryman & Bell, 2015). We use descriptive empirical evidence on the use of social media by Canadian and American public transportation agencies. In many respects, our approach mirrors the methods used by Brainard and Edlins (2015) in their study on police departments. In an era when governments and public agencies are increasing the use of social media to engage and collaborate with citizens, we want to understand whether the necessary conditions for such a collaboration exist in the first place.

What follows is a descriptive analysis of the posts and tweets from several public transportation agencies on their Facebook and Twitter accounts. Our findings will help determine the following:

Whether or not public transportation agencies use their social media accounts;

Whether or not interaction between organizations and their stakeholders and/or among stakeholders actually happens;

In the cases where interaction and dialogue are found, whether or not the latter is dialogic/collaborative in nature or oppositional/agonistic;

How conversations on social media end and the role of public transportation agencies in ending these interactions.

We analyzed the Facebook and Twitter accounts of 35 transit operators in the United States and Canada, all of which figured prominently in a report published by the TCRP in 2012. This sample represents 18 U.S. states, the District of Columbia, and five Canadian provinces (see Table 1 for details).

Table 1.

List of Public Transportation Agencies and Descriptive Statistics.

Agency name	City, State	Facebook				Twitter
Agency name	City, State	Posts	Average likes	Average sharing	Average comments	Tweets	Average favorites	Average retweet	Average replies
Bay Area Rapid Transit	Oakland, CA	21	25.19	4.10	6.38	100	10.23	11.14	1.67
Capital Area Transportation Authority	Lansing, MI	58	2.81	0.62	0.24	48	0.56	1.13	0.08
CDTA	Albany, NY	20	4.95	0.20	0.70	100	0.35	0.46	0.00
Centre Area Transportation Authority	State College, PA	32	7.09	0.56	1.41	33	0.76	0.76	0.36
Chapel Hill Transit	Chapel Hill, NC	10	1.20	0.00	0.10	10	0.90	2.60	0.20
Community Transit	Everett, WA	20	36.55	3.70	5.75	100	0.76	0.36	0.04
Dallas Area Rapid Transit	Dallas, TX	17	17.71	3.53	6.82	29	1.72	4.14	0.48
Des Moines Area Regional Transit Authority	Des Moines, IA	57	2.09	0.70	0.19	71	0.61	0.66	0.14
Edmonton Transit	Edmonton, AB	46	55.43	15.59	11.11	100	0.15	0.26	0.00
Greater Cleveland RTA	Cleveland, OH	24	5.79	0.46	2.75	100	0.37	0.60	0.05
Hillsborough Area Regional Transit	Tampa, FL	47	14.98	2.55	2.70	100	0.66	0.70	0.08
Johnson County Transit	Olathe, KS	12	1.50	0.00	0.75	50	0.40	0.38	0.22
King Country Metro Transit	Seattle, WA	21	17.71	5.14	1.81	100	0.97	1.68	0.28
Kitsap Transit	Bremerton, WA	28	2.71	1.04	0.36	NA	NA	NA	NA
LANTA	Allentown, PA	19	10.84	1.53	1.63	46	0.33	0.28	0.09
Los Angeles County Metropolitan Transportation Authority	Los Angeles, CA	100	91.63	10.54	4.17	100	2.62	2.97	0.32
Manchester Transit Authority	Manchester, NH	11	4.64	2.18	0.45	7	0.14	0.29	0.00
Metrolink	Los Angeles, CA	100	112.50	5.40	5.52	100	0.72	0.45	0.08
Metropolitan Atlanta Rapid Transit Authority	Atlanta, GA	19	9.84	1.47	1.68	100	0.14	0.28	0.02
Mountain Line Transit Authority	Morgantown, WV	5	1.20	0.00	0.20	73	0.22	0.11	0.11
MTA	New York, NY	15	247.60	62.60	32.47	88	7.16	7.55	1.45
NJ TRANSIT	Newark, NJ	30	41.43	2.17	10.27	99	1.37	3.77	1.68
Pinellas Suncoast Transit Authority	St. Petersburg, FL	100	3.38	0.15	0.15	100	0.86	0.37	0.04
Port Authority of Allegheny Country	Pittsburgh, PA	NA	NA	NA	NA	100	0.62	2.03	0.17
Potomac and Rappahannock Transportation Commission	Woodbridge, VA	19	2.42	0.05	0.26	NA	NA	NA	NA
Societe de Transport de Montreal	Montréal, QC	27	24.41	4.96	6.11	61	0.97	3.05	0.34
Sound Transit	Seattle, WA	28	47.36	8.82	7.46	100	3.72	3.62	1.26
St. Louis Metro Transit	St. Louis, MO	100	1.20	0.19	0.14	100	1.03	1.26	0.31
The Rapid	Grand Rapids, MI	80	14.49	1.91	2.88	100	0.77	0.45	0.02
Toronto Transit Commission	Toronto, ON	22	34.27	18.64	12.09	100	0.85	2.21	0.54
Translink	Vancouver, BC	83	24.47	6.04	9.46	100	0.57	0.45	0.08
Triangle Transit	Durham, NC	4	4.50	0.25	1.00	NA	NA	NA	NA
TRIMET	Portland, OR	18	213.17	98.06	21.56	100	1.36	5.15	1.37
Washington Metropolitan Area Transit Authority	Washington, DC	29	61.52	111.10	7.24	100	0.35	3.98	0.78
Winnipeg Transit System	Winnipeg, MB	NA	NA	NA	NA	100	0.17	0.24	0.43
TOTAL		1,222	35.95	8.55	4.37	2,615	1.44	2.07	0.43
STANDARD DEVIATION		30.16	57.17	26.57	6.85	29.48	2.10	2.45	0.51
MEAN		37.03	34.75	11.34	5.02	81.72	1.33	1.98	0.40
MEDIAN		24.00	14.49	2.17	2.70	100.00	0.74	0.73	0.19

Note. CDTA = Capital District Transportation Authority; RTA = Regional Transportation Authority; LANTA = Lehigh and Northampton Transportation Authority; MTA = Metropolitan Transportation Authority.

Most transportation agencies (80%) are located in large metropolitan areas, which are defined as urban areas that have a population of 200,000 or more; 17% are in small urban areas (population 50,000-200,000); and one agency (3%) is located in a rural area (TCRP, 2012). All transit modes are represented, including rail, bus, vanpool, demand-response, and ferry. Annual ridership ranges from fewer than 500,000 trips to more than 2 billion (TCRP, 2012).

We focus on Facebook and Twitter because they are the most popular social media among public transit organizations and the general public. “Facebook Quarterly Earnings Slides Q4 2015” (2016) and “Twitter Q4 2015 Company Metrics” (2016) have approximately 1.591 billion and 305 million monthly active users, respectively.

The unit of data collection for the content analysis of Facebook consists of each post generated by the organization on its official page. The unit of data collection for Twitter, meanwhile, consists of each original tweet generated by the organization. Some agencies have more than one Twitter account. For instance, Dallas Area Rapid Transit uses @dartmedia as its general account, but also uses @dartalert for alerts. Similarly, the Toronto Transit Commission uses @TTCnotices as a general account, but also @TTChelps for customer service. Whenever this situation arose, we analyzed the general account. We studied posts and tweets that were published between May 1, 2015, and June 30, 2015, with a limit of 100 posts and 100 tweets per organization.

The research team was composed of five members: two academic supervisors, a coordinator of the content analysis, and two scholars with proficiency in English. Some tests were conducted to highlight ambiguous or unclear interpretation of the coding rules based on the content of posts. The results were compared, and the differences of interpretation were discussed. This resulted in a final set of classification rules for the content of posts and tweets. Given the vast amount of data at our disposal, we decided to divide the actual content analysis among various team members. All in all, our team categorized and analyzed 1,222 Facebook posts and 2,615 tweets. We also studied the number and tone of every comment (Facebook) and reply (Twitter). Afterward, the supervisors and the coordinator compared the results obtained by the other members to ensure that there were no differences of interpretation; we obtained a Cohen’s kappa coefficient of inter-reliability of .8.

The first step of our content analysis was to categorize the content of each post/tweet in one of the following categories:

Public information (coded PI): posts or tweets providing the public with information about services, fares, long-range planning projects, and timely updates that share real-time service information and advisories with their riders (e.g., info about delays and disruptions).

Potential stakeholder and public engagement (coded SE): posts or tweets that attempt to connect with the organization’s customers, employees (e.g., recognizing current workers and recruiting new employees), and local communities in an informal way (e.g., customer suggestions for new services). For this specific category, we also assessed which type of stakeholder—shareholders/investors, employees, suppliers, customers/end users, state/public administration, third-sector organizations/NGOs, and community/general—was engaged.

Entertainment and other ancillary functions (coded O): a residual category for posts and tweets that sought to create a rapport with riders (e.g., Christmas greetings or online prize competitions) or entertain them through songs, videos, and contests.

Table 2 shows an example of a Facebook post that features potential stakeholder and public engagement. Table 3 shows a tweet that contains public information on services. Table 4 shows an example of a Facebook post in the residual category (texts are reported as is, including typos, but user identities and pictures have been removed for anonymity).

Table 2.

Transcript of a Sample Facebook Post of Potential Stakeholder and Public Engagement.

Post published by the agency
Metro Forward
June 12, 2014 [Post’s privacy set to “Public”]
[Photo of a smiling driver]
Metro is looking to hire over 200 Bus Operators this summer! Looking for a career with growth opportunities? This is the position for you.
Minimum qualifications: high school diploma or GED; valid driver’s license; 3 years of post-HS employment or 3 years of full/part-time student status & employment history. Visit [URL] and search for job code 150268. #wmatajobs #wmatacareers
[469 likes] [128 comments] [3,004 shares]
Examples of comment	Examples of replies by the agency	Examples of replies by other Facebook users
I got my CDL permit but not the license. Does Metro provide training to obtain my license? [4 likes, 11 replies]	[Metro Forward] Yes, we do assist CDL permit holders with obtaining their licenses. [7 likes]	Last I heard a couple of months ago they couldn’t . . . . . Lost the license to train. My barber work for metro [2 likes]
		I applied a few months ago and received a email stating what [another user] said . . . [0 likes]
Someone told me metro don’t hire felons is this true???? [3 likes, 0 replies]	—	It depends on your charge, some felonies are except if they are over 10 years and some are indefinite. They will tell you what charges will disqualify you before you take the test. [0 likes]

Note. GED = General Education Development; CDL = Commercial Driver’s License.

Table 3.

Transcript of a Sample Tweet Containing Public Information.

Post
@SFBART [Bay Area Rapid Transport]
June 24, 2015
[Map of the disrupted lines]
No train service between SF & East Bay during 2 upcoming weekends. [link to a page with more details]
[37 favourites] [22 replies] [120 retweet]
Examples of replies	Examples of replies by the agency	Examples of replies by other Twitter users
@SFBART Please be more specific about the affected weekends. [1 reply]	@[username] Thanks for asking. Attached is weekend service details for Aug 1-2, Sep 5-7 at http://ow.ly/d/3rpmorhttp://ow.ly/OKOL4	—
@SFBART @[other username] So presumably, people headed to WO would have to make their own way from 19th st? No bus bridge there? [1 reply]	@[username] Yes, #SFBART bus bridge is 19th St station to temporary Transbay Terminal. Riders can call 510-465-BART for @rideact options	—
@SFBART Thanks for #tweeting this alert #sfbart! #sanfrancisco #sfbayarea #publictransit #oakland #eastbay #trains #infrastructure #repairs [0 replies]	—	—

Table 4.

Transcript of a Sample Facebook Post for the Residual Category (Entertainment and Other Ancillary Functions).

Post published by the agency
Toronto Transit Commission
June 24, 2014 [Post’s privacy set to “Public”]
TTC riders, blogTO readers to decide their favourite transit app. The Toronto Transit Commission and blogTO would like everyone to exercise their democratic right and vote . . . for their favourite transit app.
The jointly-sponsored contest is now live on blogTO. Earlier this year, blogTO and the TTC began the process of identifying candidates best-suited for both iPhone and Android users. The top three nominees are (in alphabetical order):
MOOVIT
ROCKETMAN
TRANSITAPP
The public may cast their vote at [URL]. The last day of voting is July 13. The winning app, to be announced on July 14, will be promoted by the TTC on its website as the TTC/blogTO transit app winner.
Last December, former TTC Chair Karen Stintz proposed that the TTC undertake a transit app contest to help improve customer knowledge of self-serve opportunities with third-party mobile apps using TTC open data.
[2 likes] [3 comments] [0 shares]
Examples of comment	Examples of replies by the agency	Examples of replies by other Facebook users
People are over an hour late for work today—they don’t care about your app if your actual service doesn’t work. Post about THOSE updates. For the last half hour the news hasn’t had ANY updates—work on THAT. [0 likes, 2 replies]	—	I disagree. [0 likes]
	—	[REJOINDER] You can disagree tomorrow—I’m talking about how inappropriate it is to be posting this today when there are far greater issues. [0 likes]

Note. TTC = Toronto Transit Commission uses.

Second, we assessed what level and what type of interaction was effectively reached for every category of post and tweet. We collected the number of likes, shares, and comments (for Facebook accounts) and favorites, retweets, and replies (for Twitter accounts) for every single unit in each category. This is relevant because a high number of likes/favorites, sharing/retweets, and replies/comments suggests a reasonable level of interaction between agencies and users.

Because we want to assess the contribution of social media to stakeholder engagement (SE) in a system of dialogic accounting, we also tracked whether (and how) agencies reply to comments on their posts, thereby creating a true two-way conversation with users, and whether users communicate with each other on the organization’s Facebook or Twitter page. To do this, we defined two “interaction ratios.” The first interaction ratio—“Interaction ratio (agency)”—varies between 1, when the organization replied to at least one comment (Facebook) or replied (Twitter) in every post, and 0, when the organization did not reply to any comment/tweet. The second interaction ratio—“Interaction ratio (users)”—varies between 1, when at least one user replied to at least one other user comment (Facebook) or replied (Twitter) in every post, and 0, when users did not reply to any comment/tweet. This analysis was carried out to understand how often local transportation agencies take part in two-way discussions and assess the extent to which users interact with each other.

In addition, we also investigated every comment/reply to assess the general tenor of the discussion. In particular, we wanted to determine whether feedback from users was more positively oriented (compliments for the agency’s activities, constructive proposals, etc.), negatively oriented (protests, constructive criticism, etc.), or neutral. In this fourth step, we classified each comment/reply with a value ranging from 1 (protest) to 5 (compliment). Comments with constructive criticism received a value of 2, constructive proposals received a value of 4, and neutral comments received a value of 3. We then proceeded to build an index for every category of units that showed the average tone of comments, using a scale of 1 to 5. All of the data were collected into two databases, one for Facebook posts and one for Twitter tweets. Spam and off-topic comments were excluded from the analysis.

Results

We determined that 33 transportation agencies out of 35 have an active Facebook page that features at least one post in the period of our analysis, and 32 out of 35 have an active Twitter account. Table 1 shows the amount of posts and tweets published by each operator on its official Facebook page and Twitter account between May 1, 2015, and June 30, 2015 (with a maximum of 100 posts and 100 tweets per organization). Table 1 also shows the average number of likes, shares, and comments on the Facebook posts included in our analysis, as well as the average number of favorites, retweets, and replies to tweets. Because the first part of our analysis focused on determining whether social media are used as a mechanism of public information and potential stakeholder engagement, we categorized the posts according to their aim and content. Table 5 shows the results of the content analysis on a total of 1,222 posts and 2,615 tweets.

Table 5.

Number of Posts and Tweets Categorized by Content for Each Agency.

Agency name	Type of posts							Total posts	Type of tweets									Total tweets
		If SE, which stakeholder:								If SE, which stakeholder:
	SE	Community/General	Customer/End Users	Employees	Shareholders/Investors	PI	O		SE	Community/General	Customers/End users	Employees	Shareholders/Investors	State/Public administration	Suppliers	PI	O
Bay Area Rapid Transit	4				4	5	12	21	SE	6		4		2			66	28	100
Capital Area Transportation Authority	10	8	2			24	24	58	4	4	0					28	16	48
CDTA	3	1	2			8	9	20	15	9	5	1				46	39	100
Centre Area Transportation Authority	11	5	4	2		9	12	32	15	6	4	5				7	11	33
Chapel Hill Transit	2		1	1		6	2	10	2		1	1				6	2	10
Community Transit	3		2	1		8	9	20	65	10	49	5	1			15	20	100
Dallas Area Rapid Transit	3	3	0			5	9	17	8	2	6					8	13	29
Des Moines Area Regional Transit Authority	18	9	6	1	2	18	21	57	12	1	9		2			23	36	71
Edmonton Transit	8	2	5		1	18	20	46			0					92	8	100
Greater Cleveland RTA	7	2	5			7	10	24	9	2	7					82	9	100
Hillsborough Area Regional Transit	13	4	1	2	6	20	14	47	11	5	3	1	2			54	35	100
Johnson County Transit	2		1		1	6	4	12	2		0	1	1			27	21	50
King Country Metro Transit	10	4	2	4		7	4	21	14	7	5	2				71	15	100
Kitsap Transit	11	7	3	1		14	3	28
LANTA	1		1			7	11	19	6	1	3	2				16	24	46
Los Angeles County Metropolitan Transportation Authority	13	10	3			32	55	100	6		5		1			79	15	100
Manchester Transit Authority	1		0	1		10		11			0					7		7
Metrolink			0			57	43	100	1		0		1			97	2	100
Metropolitan Atlanta Rapid Transit Authority	1	1	0			12	6	19	6		6					94		100
Mountain Line Transit Authority	3	1	1	1		1	1	5	19	5	11	2	1			37	17	73
MTA	4		4			3	8	15	24	16	6		1		1	20	44	88
NJ TRANSIT	17	5	12			5	8	30	19	1	17	1				77	3	99
Pinellas Suncoast Transit Authority	51	36	8	6	1	4	45	100	30	14	3	13				16	54	100
Port Authority of Allegheny Country									13		11		2			84	3	100
Potomac and Rappahannock Transportation Commission	11	2	9			5	3	19
Societe de Transport de Montreal	12	7	5			2	13	27	10	4	6					49	2	61
Sound Transit	11	4	7			3	14	28	24	13	11					34	42	100
St. Louis Metro Transit	19	17	2			53	28	100	9	5	3			1		67	24	100
The Rapid	24	8	8	7	1	20	36	80	33	15	10	8				14	53	100
Toronto Transit Commission	6		6			12	4	22			0					100		100
Translink	15	4	5		6	27	41	83	6		6					92	2	100
Triangle Transit	1		0		1	1	2	4
TRIMET	5	1	2	2		3	10	18			0					99	1	100
Washington Metropolitan Area Transit Authority	4		3	1		15	10	29			0					100		100
Winnipeg Transit System									83		83					14	3	100
TOTAL	304	141	110	30	23	427	491	1,222	452	120	274	42	14	1	1	1,621	542	2615
PERCENTAGE	24.8	11.54	9.00	2.45	1.88	34.94	40.18	100	17.28	4.59	10.48	1.61	0.54	0.04	0.04	61.99	20.73	100
MEAN	9.50	6.41	3.44	2.31	2.56	12.94	15.34	37.03	16.74	6.67	8.56	3.50	1.40	1.00	1.00	50.66	19.36	81.72
MEDIAN	7.50	4.00	2.50	1.00	1.00	8.00	10.00	24.00	11.00	5.00	5.00	2.00	1.00	1.00	1.00	47.50	15.50	100.00

Note. PI = public information; CDTA = Capital District Transportation Authority; RTA = Regional Transportation Authority; LANTA = Lehigh and Northampton Transportation Authority; MTA = Metropolitan Transportation Authority.

To answer our exploratory research question, we need to understand what kind of content transportation agencies deliver on social media. During the period of analysis, organizations in our sample published 304 Facebook posts (24.88% of the total Facebook posts) with the aim of engaging one or more of their stakeholders (SE posts). Four hundred twenty-seven Facebook posts (34.94%) provided public information (PI posts), including information about services, real-time updates, or disruptions. SE and PI posts on Facebook accounted for 59.82% of total posts; posts on entertainment and other ancillary functions (O) accounted for the remaining 40.18%. In terms of Twitter usage, 452 tweets (17.28%) were aimed at potential stakeholder engagement and 1,621 (61.99%) were aimed at issuing public information. SE and PI tweets accounted for 79.27% of total tweets, while tweets labeled 0 accounted for the remaining 20.73%. Twitter appears to be more focused on PI and SE than O. These results show that public transportation agencies often take advantage of their presence on social media to provide the public with information on their services and perform activities involving potential stakeholder engagement. As Table 5 shows, posts and tweets aiming at potential stakeholder engagement are generally targeted toward customers, end users, or the community as a whole.

By comparing our findings with other studies in this field (see, for example, Brainard & Edlins, 2015), we argue that public transportation agencies produced medium- or high-activity levels on their social media pages, which was probably due to the specific features of their industry (e.g., close relationship with users, direct service provisions, and the importance of keeping users informed). However, our results show some significant differences in terms of how these two social media are utilized. Twitter is mainly used for PI messages, especially short, real-time updates on services, delays, or disruptions. Both Facebook and Twitter are used for SE activities, but the former appears to be used more than the latter to interact with stakeholders in a dialogic perspective. Standard deviation values confirm that there are differences in the sample in terms of how public transportation agencies approach social media. Tables 1 and 5 show mean and median values to provide the reader with further guidance on how we interpreted the different levels of activity on each organization’s social media page. We can confirm that these types of organizations, with a few exceptions, have established an active presence on Facebook and Twitter, and regularly publish new content. Although the willingness to engage stakeholders may vary across the sample, nearly all of the organizations published at least one Facebook post on SE and one tweet on PI in the period of analysis.

This trend is confirmed by data on online interaction and conversations collected during the second and third steps of our content analysis. Table 6, which shows social media data on interaction through Facebook posts (Panel A) and tweets (Panel B), illustrates the extent to which Facebook produced a higher grade of interaction. The total and average values of likes, shares, and replies are notably higher on Facebook, thus confirming the notion that users are less likely to interact with public transportation agencies on Twitter. An average value of 24.67 likes, 17.66 shares, and 4.98 comments were found in Facebook posts that were categorized as SE. Tweets devoted to SE, meanwhile, collected an average value of 1.07 favorites, 1.21 retweets, and 0.38 replies. This ratio is roughly the same for PI messages: Facebook has an average value of 25.91 likes, 4.47 shares, and 3.94 comments, while Twitter received an average value of 1.01 favorites, 2.23 retweets, and 0.44 replies. Facebook posts categorized as O received an average value of 51.67 likes, 6.46 shares, and 4.36 comments; tweets in the same category collected an average value of 3.02 favorites, 2.29 retweets, and 0.46 replies. Posts and tweets on entertainment and other ancillary functions (O) present the highest level of likes and favorites. These findings remain significant even if we take into account the fact that Facebook has roughly 5 times the active monthly users of Twitter. In fact, even using this term of confrontation, the interaction data remain much higher on Facebook than on Twitter.

Table 6.

Social Media Data on Interaction by Type of Facebook Post and Tweet.

Panel A—Facebook
Category	No. of posts	Total no. of likes	Likes (M)	Total no. of shares	Shares (M)	Total no. of comments	Comments (M)
SE	304	7,499	24.67	5,368	17.66	1513	4.98
PI	427	11063	25.91	1,908	4.47	1683	3.94
O	491	25370	51.67	3,174	6.46	2139	4.36
TOTAL	1,222	43,932	35.95	10,450	8.55	5335	4.37
SD.	95.04	9,458.37	15.24	1,750.62	7.11	323.71	0.52
Panel B—Twitter
Category	No. of tweets	Total no. of favorites	Favorites (M)	Total no. of retweets	Retweets (mean)	Total no. of replies	Replies (mean)
SE	452	483	1.07	545	1.21	173	0.38
PI	1,621	1,636	1.01	3,616	2.23	713	0.44
O	542	1,639	3.02	1,242	2.29	248	0.46
TOTAL	2,615	3,758	1.44	5,403	2.07	1134	0.43
SD	650.50	666.55	1.15	1,610.01	0.61	292.53	0.04

Note. PI = public information.

We can conclude, then, that interaction is higher on Facebook and is more likely to feature content that tries to engage stakeholders. Conversely, Twitter tends to focus on messages containing PI. We can affirm that public transportation agencies often use social media to engage stakeholders and publish public information, albeit in ways that vary according to which type of social medium is being used.

The behaviors of users and organizations are different on Facebook and Twitter, especially if we take levels of interaction into account. Our content analysis assessed the level and type of interaction between public transportation agencies and users for each category of posts on Facebook and Twitter. In the third and fourth steps of our analysis, we also investigated the tenor of these comments and the extent to which the organizations in our sample replied to user comments. This is important because it helps us understand whether social media actually support a system of dialogic accounting for public transportation agencies, and whether they lead to two-way conversations between organizations and their stakeholders. Table 7 shows the number of Facebook posts with comments (Panel A) and tweets with replies (Panel B), two indexes representing the level of interaction between agencies and users and among users, and an index describing the tenor of comments.

Table 7.

Level of Interaction and Tenor of Comments/Replies for Facebook and Twitter.

Panel A: Facebook
Category	No. of posts	No. of posts with comments	No. of posts with replies from other users	No. of posts with replies from the agency	Interaction index (users)	Interaction index (agency)	Tenor of comments (index)
SE	304	140	65	60	0.464	0.429	2.829
PI	427	203	90	86	0.443	0.424	2.724
O	491	274	113	85	0.412	0.310	3.172
TOTAL	1,222	617	268	231	0.434	0.374	2.947
SD	95.04	67.04	24.01	14.73	0.03	0.07	0.23
Panel B—Twitter
Category	No. of tweets	No. of tweets with replies	No. of tweets with replies from other users	No. of tweets with replies from the agency	Interaction index (users)	Interaction index (agency)	Tenor of replies (index)
SE	452	78	6	24	0.077	0.308	2.897
PI	1,621	268	18	62	0.067	0.231	2.306
O	542	104	9	18	0.087	0.173	2.808
TOTAL	2,615	450	33	104	0.073	0.231	2.524
SD	650.50	103.01	6.24	23.86	0.01	0.07	0.32

Note. PI = public information.

Because we intend to examine the contribution of social media to SE and PI in a framework of dialogic accounting, we also studied whether public transportation agencies create two-way conversations with users and whether users communicate with each other. The interaction ratios presented in Table 7 indicate whether (and how) organizations reply to comments on their official accounts. On Facebook, SE posts present a “users interaction index” of 0.464 (meaning that users replied to other users’ comments in 46% of cases) and an “agency interaction index” of 0.429 (meaning that agencies replied to at least one user’s comment in 43% of posts). PI posts, meanwhile, feature an interaction index for users of 0.443 and an interaction index for agencies of 0.424. These values confirm that interaction on Facebook is higher than on Twitter and that users and agencies often engage in two-way conversations. In fact, organizations in our sample effectively took part in conversations—answering users’ contributions—in nearly half of posts that were published with the aim of engaging users and stakeholders.

The interaction dynamic is completely different on Twitter, where SE tweets present an interaction index for users of 0.077 and an interaction index for agencies of 0.308. Similarly, PI tweets feature an interaction index for users of 0.067 and an interaction index for agencies of 0.231. The interaction within users is much lower on Twitter than on Facebook. It is interesting to note, however, that the total agency interaction index (0.231) is considerably higher than the total user interaction index (0.073), thereby showing that public transportation agencies respond to user comments and start a conversation in 23% of cases. This number rises to 30% when we look exclusively at tweets that are geared toward SE. Indeed, SE-related content tended to produce higher interaction values on both Twitter and Facebook. Although PI posts—for example, news on disruptions or new forms of payments—can lead to a high level of interaction, our data confirm that users react with a higher degree of intensity to content that seeks to engage stakeholders by requesting an action or an opinion.

We also wanted to determine whether feedback from users is positively oriented, negatively oriented, or neutral. To assess the general tenor of the discussion, we examined the contents and tone of each Facebook comment or Twitter reply and built an index—ranging from 1 (critical) and 5 (supportive)—following the procedure described in the “Methodology” section of this article. We found that the average tenor of comments to posts published by public transportation agencies is mildly negative. As Table 7 shows, Facebook posts related to SE and PI present a value of 2.829 and 2.724; tweets related to SE and PI, meanwhile, feature a value of 2.897 and 2.306. We noted a prevalence of critical comments—especially on Twitter—with regard to the general quality of the services, lines disruptions, or delays. Facebook posts on amusement and other ancillary functions (O) present the best tenor of comments index (3.172).

An analysis of the values in this index can help us understand whether the interaction produced by these organizations is generating discussions with a generic consensus or whether an agonistic perspective, which highlights struggles and differences between actors, prevails. This index does not represent a direct indicator of either deliberative consensus and democratic engagement or divergent and agonistic perspectives. This is because the information we collected is just a “snapshot” of extemporaneous online conversations. It can, however, suggest an orientation toward one of these two forms of dialogic communication, although more in-depth, case by case studies need to be carried out. Other factors—for instance, the specific dynamics found in interactions between users and their reception by organizations—deserve special attention.

Because the median value of 3 indicates a neutral/mixed tenor, we interpret average values of the tenor of comments index lower than 3 as a possible evidence of potential agonistic approach. However, we believe it is necessary to assume a cautious position, as we are dealing with agencies delivering critical local transportation services whose activities are likely to result in significant expressions of dissent, especially in the case of conversations involving disruptions and delays.

Moreover, Table 8 provides an overview of every interaction value—average number of likes/favorites, comments/replies, sharings/retweets, comments/replies by users to users, comments/replies by agencies, and the tenor of comments/replies—for each agency and compares them with average weekday ridership (in 000s). Our findings suggest a positive correlation between the size of an organization and the average number of likes/favorites, sharings/retweets, and comments/replies. Indeed, the Pearson correlation coefficient between the number of Facebook comments and the size of the organization is highly positive (.811). There is no correlation, however, between size and the number of comments/replies by agencies who use either Facebook or Twitter. The average number of comments by users to other users on Facebook is found to be highly correlated to size (.818), while the number of user replies to other users on Twitter is only mildly correlated. Finally, the tenor of the discussion is negatively correlated to the size of the organization on both Facebook (−.459) and Twitter (.407). Transportation agencies operating in smaller communities usually received more appreciative posts and tweets (low interaction index but higher tenor index). Agencies serving bigger communities, conversely, face more mixed feedback, and conversations are both more interactive and less focused. In fact, it is quite common to see heated discussions on the pages of bigger organizations, as well as more instances of spam. Table 9 presents an overview of these correlation coefficients.

Table 8.

Average Weekday Ridership (in 000s) and Social Media Interaction Values for Each Agency.

Agency name	Avg. daily ridership (000s)	Facebook						Twitter
Agency name	Avg. daily ridership (000s)	Avg. Likes	Avg. Shar.	Avg. Comm.	Avg. comm. Agency	Avg. comm. Users	Tenor comm.	Avg. Fav.	Avg. Retw.	Avg. Replies	Avg. repl. Agency	Avg. repl. Users	Tenor comm.
Bay Area Rapid Transit	458.9	25.19	4.10	6.38	0.00	2.14	2.35	10.23	11.14	1.67	0.11	0.06	2.66
Capital Area Transportation Authority	30.9	2.81	0.62	0.24	0.00	0.00	3.25	0.56	1.13	0.08	0.00	0.00	4.00
CDTA	55.4	4.95	0.20	0.70	0.05	0.00	3.75	0.35	0.46	0.00	0.00	0.00	NA
Centre Area Transportation Authority	17.2	7.09	0.56	1.41	0.19	0.34	3.50	0.76	0.76	0.36	0.18	0.00	4.13
Chapel Hill Transit	23.3	1.20	0.00	0.10	0.00	0.00	2.00	0.90	2.60	0.20	0.10	0.00	5.00
Community Transit	34.6	36.55	3.70	5.75	0.65	1.15	3.93	0.76	0.36	0.04	0.01	0.00	3.67
Dallas Area Rapid Transit	230.3	17.71	3.53	6.82	1.47	1.06	1.92	1.72	4.14	0.48	0.14	0.00	3.50
Des Moines Area Regional Transit Authority	20.8	2.09	0.70	0.19	0.11	0.04	3.29	0.61	0.66	0.14	0.07	0.03	3.00
Edmonton Transit	NA	55.43	15.59	11.11	0.96	6.65	3.14	0.15	0.26	0.00	0.00	0.00	NA
Greater Cleveland RTA	NA	5.79	0.46	2.75	0.54	0.25	2.20	0.37	0.60	0.05	0.01	0.01	1.25
Hillsborough Area Regional Transit	48.1	14.98	2.55	2.70	0.38	0.49	3.16	0.66	0.70	0.08	0.02	0.01	3.33
Johnson County Transit	2.2	1.50	0.00	0.75	0.25	0.00	3.00	0.40	0.38	0.22	0.04	0.00	3.00
King Country Metro Transit	410.9	17.71	5.14	1.81	0.05	0.33	3.22	0.97	1.68	0.28	0.05	0.01	2.19
Kitsap Transit	14.4	2.71	1.04	0.36	0.07	0.00	2.80	NA	NA	NA	NA	NA	NA
LANTA	21.4	10.84	1.53	1.63	0.16	0.00	2.89	0.33	0.28	0.09	0.04	0.00	4.00
Los Angeles County Metropolitan Transportation Authority	1,371.2	91.63	10.54	4.17	0.08	1.12	3.12	2.62	2.97	0.32	0.07	0.05	2.87
Manchester Transit Authority	2.5^a	4.64	2.18	0.45	0.09	0.09	3.00	0.14	0.29	0.00	0.00	0.00	NA
Metrolink	40.8	112.50	5.40	5.52	0.43	0.71	3.16	0.72	0.45	0.08	0.00	0.00	1.83
Metropolitan Atlanta Rapid Transit Authority	451.6	9.84	1.47	1.68	0.16	0.42	2.36	0.14	0.28	0.02	0.00	0.00	3.00
Mountain Line Transit Authority	1.3^a	1.20	0.00	0.20	0.00	0.00	5.00	0.22	0.11	0.11	0.04	0.00	3.50
MTA	11,097.40	247.60	62.60	32.47	0.00	8.87	1.60	7.16	7.55	1.45	0.01	0.08	1.90
NJ TRANSIT	NA	41.43	2.17	10.27	0.27	1.53	1.43	1.37	3.77	1.68	0.15	0.06	1.81
Pinellas Suncoast Transit Authority	45.5	3.38	0.15	0.15	0.04	0.01	3.25	0.86	0.37	0.04	0.04	0.03	3.25
Port Authority of Allegheny Country	NA	NA	NA	NA	NA	NA	NA	0.62	2.03	0.17	0.00	0.07	3.00
Potomac and Rappahannock Transportation Commission	10.6^a	2.42	0.05	0.26	0.00	0.00	5.00	NA	NA	NA	NA	NA	NA
Societe de Transport de Montreal	1,858	24.41	4.96	6.11	1.85	1.37	2.95	0.97	3.05	0.34	0.07	0.00	2.86
Sound Transit	121	47.36	8.82	7.46	0.39	1.71	3.18	3.72	3.62	1.26	0.17	0.16	3.31
St. Louis Metro Transit	145.1	1.20	0.19	0.14	0.08	0.01	3.30	1.03	1.26	0.31	0.11	0.07	3.12
The Rapid	37.1	14.49	1.91	2.88	0.34	0.49	3.11	0.77	0.45	0.02	0.00	0.00	4.00
Toronto Transit Commission	2,569.3	34.27	18.64	12.09	0.23	3.41	2.00	0.85	2.21	0.54	0.26	0.00	2.61
Translink	1,171.9	24.47	6.04	9.46	2.20	5.11	3.10	0.57	0.45	0.08	0.09	0.08	2.60
Triangle Transit	NA	4.50	0.25	1.00	0.00	0.00	3.00	NA	NA	NA	NA	NA	NA
TRIMET	321.9	213.17	98.06	21.56	3.89	4.67	3.28	1.36	5.15	1.37	0.23	0.10	1.81
Washington Metropolitan Area Transit Authority	1,343.6	61.52	111.10	7.24	0.41	2.59	2.85	0.35	3.98	0.78	0.02	0.01	1.80
Winnipeg Transit System	130^a	NA	NA	NA	NA	NA	NA	0.17	0.24	0.43	0.16	0.03	2.90
AGGREGATE	736.24	35.95	8.55	4.37	0.46	1.31	2.95	1.44	2.07	0.43	0.07	0.03	2.52
STANDARD DEVIATION	2,057.87	57.17	26.57	6.85	0.81	2.13	0.78	2.10	2.45	0.51	0.07	0.04	0.86

Note. Ridership data are provided by American Public Transportation Association (2015). CDTA = Capital District Transportation Authority; RTA = Regional Transportation Authority; LANTA = Lehigh and Northampton Transportation Authority; MTA = Metropolitan Transportation Authority.

Indicates values provided directly by the transportation agency.

Table 9.

Pearson Correlation Coefficients Between Interaction Values and Agency Size.

Correlation with size	Avg. Likes/Fav.	Avg. Shar./Retweets	Avg. Comm./Replies	Avg. Comm./Replies Agency	Avg. Comm./Replies Users	Tenor of Comm./Replies
Facebook	.690	.419	.811	−.023	.818	−.459
Twitter	.498	.475	.472	−.050	.257	−.407

Conclusion

The present research examined whether online interaction through social media is used as a mechanism of public information and potential stakeholder engagement by public transportation agencies. Using dialogic accounting theory as a point of reference, we conducted a content analysis on the Facebook and Twitter pages of 35 Canadian and American transit agencies.

Social media technology represents a new way for organizations and public agencies to connect and collaborate with their users and stakeholders (Brainard & Edlins, 2015). Social media are often used by local governments in hopes of improving the delivery of services to citizens and encouraging civic engagement and participation (Holzer & Manoharan, 2008). Although it is important to understand the use of technology to improve efficiency in public organizations, it is increasingly important to understand the ways in which different types of digital technologies are shaping government behavior and why some governments adopt some technologies and not others (Li & Feeney, 2014).

Building on these premises, our results confirm that many public transportation agencies have an active presence on Facebook and Twitter. The results of our content analysis show that organizations in our sample often take advantage of social media to provide public information on their services and engage stakeholders. However, our analyses also show some significant differences in how Facebook and Twitter are used. Twitter, for instance, is most often used for PI messages, especially short, real-time updates on services, delays, or disruptions. Both social media are used for messages engaging stakeholders, but Facebook appears to be used more than Twitter to publish content aiming at interacting in a dialogic perspective and creating two-way conversations with users. This result is consistent with the findings of Brainard and Edlins (2015). They claim that Facebook appears to be the most promising technology for dialogue between local authorities and residents. Twitter, they argue, may be negatively affected by the 140-character limit on individual tweets. Facebook is structured to provide “deeper” content than Twitter; this, together with its broader user base, suggests that Facebook is a more robust tool in which to engage stakeholders in a two-way conversation. Although our findings confirm this assumption, it should be noted that Twitter also has several features that make it an ideal tool in which to interact with users. Because Twitter provides users with a platform in which to quickly publish well-organized, user-oriented messages (e.g., news on service provisions), one could also argue that it provides users with the means of reacting and providing feedback in a similarly efficient manner. We believe our findings provide a clear picture of the strategies public transportation agencies have adopted with regard to various social media applications.

Social media might help us embrace more dialogic forms of communication, a notion that is supported by much of the data in the current study. For instance, the higher total and average values of likes, shares, and replies on Facebook confirm that users prefer to interact with public transportation agencies on this social medium than on Twitter. Interaction is higher and focused on Facebook, especially with regard to content that tries to engage stakeholders. Twitter, meanwhile, features lower levels of interaction and is most often focused on public utility information. If we take the levels of interaction into account, we can then confirm that the behavior of both public transportation agencies and the social media users who visit their pages are different on Facebook and Twitter, as both take advantage of the peculiar characteristics of each type of social medium. The practical implications of our findings are clear. For instance, public transportation agencies that do not have a clear strategy on how to use social media might feel impelled to focus on Facebook as a means of achieving a better form of stakeholder engagement, while Twitter could be used to instantaneously disseminate information as a means of avoiding misunderstandings among users.

Our findings regarding interaction levels show that users and agencies are often actively engaged in two-way conversations on Facebook, as public transportation agencies often reply to users and users often reply to each other. This led us to conclude that creating enduring two-way conversations on social media can encourage stakeholder engagement and dialogic communication, while also supporting various forms of dialogic accounting, thereby assisting organizations in their attempts to collect relevant information from various stakeholders. However, social media can also be used to support more traditional management accounting tools—particularly those that assist management in the formulation and implementation of an organization’s strategy—because they help in the collection of more and better information from various stakeholders. In fact, social media have the potential to support forms of dialogic accounting because they can help better define materiality and relevance of information.

Nonetheless, we believe that social media’s potential has not yet been fully exploited for at least two reasons. First, many organizations are not fully aware of the diversity of stakeholder values and interests. In other words, our results confirm that organizations are slow to adopt a pluralist vision that might expose them to divergent values and interests. In other words, one can interpret this lack of commitment as a fear that their reputation and legitimacy will be damaged. Although it is relatively easy to exploit and manage the legitimizing effects of social media, organizations risk exposing themselves to the effects of negative feedback if these tools are used in a mono-directional way. Suffice to say that organizations often prefer to reduce the opportunities for these kinds of comments instead of addressing them directly.

Second, and perhaps most importantly, the emergence of social media is a relatively new phenomenon and organizations that serve both small and large communities need to learn the best way to manage them, especially if they want to use them for something more than simply promoting a service or a product. Special attention is needed to examining how staff dedicated to social media tend to respond to comments and replies by users.

Brainard and Edlins (2015) argue that although local authorities and citizens communicate with each other on social media, interaction levels could be expanded if local authorities were more pro-active. Our findings are consistent with these conclusions, suggesting that organizations should place more emphasis on engaging users and citizens in dialogic conversations. Publishing engaging content on social media and providing a public space for comments is not enough; it is necessary to expand on these discussions to encourage a more meaningful brand of dialogic communication. We therefore conclude that better management of these conversations will result in a more collaborative process.

Our analyses confirm that the number of answers to users on social media is not necessarily related to the size of the agency itself. We have found, based on our exploratory studies, that local transportation agencies (and public agencies in general) could create more interactive and fruitful conversation without having a significant impact on their budget. Indeed, public agencies could still be selective in answering questions posed by users while also addressing constructive critiques without fear of damaging their reputation or legitimacy. Further studies ought to analyze the impact of a dedicated social media staff on the budget and quality of services of public agencies.

The results of our study on the tenor of Facebook and Twitter comments help us understand whether the interaction encouraged by public transportation agencies leads to the creation of consensus, or whether it encourages agonistic approaches that highlight the struggles and differences between actors. The data we collected on Facebook and Twitter led us to argue in favor of the latter. This is not surprising, as we are dealing with organizations that deliver critical transportation services and whose actions are likely to result in significant expressions of dissent, especially because the majority of conversations revolve around service disruptions or delays.

Nonetheless, there are some significant differences between the types of comments found on Facebook and those found on Twitter. Given the different features of Facebook and Twitter, and the different types of content posted on each social network, our results show that Facebook presents some features of a more deliberative approach and Twitter presents some features of a more agonistic perspective. However, as noted in the “Results” section of this article, the information we collected is just a “snapshot” of extemporaneous online conversations; accordingly, we can only confirm that some preliminary evidence of divergent conversations emerge, although more robust evidence is needed.

Furthermore, Facebook is more flexible and more capable of giving space to complex commentary, while Twitter enables rapid and short responses. Our results also show that Facebook pages are sometimes used as public arenas that sustain conversations, focusing on the interaction not only between organizations and stakeholders but also between users and stakeholders themselves. Interaction values confirm that Facebook users often show a willingness to interact with other users, oftentimes to help other users answer a question or confirm a complaint by other customers. As a result, transportation agencies are given the opportunity to observe and analyze user behavior and make any necessary adjustments.

The current study has at least three limitations that ought to be addressed in future studies. First, and most importantly, further research should address the outcomes of potential stakeholder engagement and, in particular, how interactions through social media affect the quality of public services. In light of our results, we can only confirm that preliminary evidence of stakeholder engagement exists on Facebook, although this does not necessarily imply that effective stakeholder engagement has been conducted. Further research on this topic should examine other forms of evidence and move beyond mere content analysis. For instance, adopting more qualitative approaches, including interviews with both representatives of public transportation agencies and users, minutes of participatory meetings, and archival analysis, could be especially fruitful.

Second, future research should make an effort to provide in-depth case studies on how organizations invest in social media interaction, manage their profiles and mobile apps, and address the interaction between users.

Third, because this study focuses solely on Canadian and American organizations, future research should expand into different geographical areas.

In terms of its practical implications, this study helps shed light on the utilization of social media by public transportations agencies, thereby drawing more attention to how innovative technologies can be used to engage users and stakeholders in both the public and private sectors. Our study suggests that there is room for improvement in terms of how public agencies manage social media as a means of supporting dialogic accounting systems. We found that many comments (on Facebook) and replies (on Twitter) were left unanswered by public agencies. This represents a missed opportunity to create two-way conversation with users. Addressing all of these comments or replies would encourage a more effective use of social media in terms of supporting stakeholder engagement. Perhaps more importantly, we believe that unanswered comments or replies only serve to discourage users from interacting with these organizations in the future. We believe that although a broader and more continuous commitment to social media interaction between users and stakeholders will mildly affect an organization’s budget, it will also create new opportunities for improving its services.

Footnotes

Acknowledgements

The authors would like to thank the anonymous reviewers whose constructive suggestions helped improve the article and the team of junior researchers—in particular, Arianna Bini and Cora Dormendo—who supported the phase of data collection and data entry of this study.

Authors’ Note

Although the article is the result of a team effort, Giacomo Manetti can be considered the author of the “Introduction” and “Literature Review” sections, and Marco Bellucci the author of the “Methodology” and “Results” sections. The “Conclusion” section has been written by the three authors.

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.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Giacomo Manetti (PhD) is an associate professor of accounting at the Department of Economics and Management of the University of Florence (Italy). His research interests include stakeholder theory, corporate social responsibility, sustainability reporting and assurance, third sector, and accounting history.

Marco Bellucci is a PhD student in business administration and management at the University of Florence. His research interests include corporate social responsibility, sustainability reporting, third-sector organizations, and social enterprises.

Luca Bagnoli (PhD) is a full professor of management accounting for cooperatives and nonprofit organizations at the Department of Economics and Management of the University of Florence (Italy). His research interests are social enterprises, cooperatives, and voluntary organizations.

References

Adams

(2002). Internal organisational influencing corporate social and ethical reporting: Beyond current theorizing. Accounting, Auditing & Accountability Journal, 15, 223-250.

American Public Transportation Association. (2015). Transit ridership report Third Quarter 2015. Washington, DC: American Public Transportation Association.

Andriof

Waddock

Husted

Rahman

(2002). Unfolding stakeholder thinking: Theory responsibility and engagement. Sheffield, UK: Greenleaf Publishing.

Ballou

Heitger

D. L.

Landes

C. E.

(2006). The future of corporate sustainability reporting: A rapidly growing assurance opportunity. Journal of Accountancy, 22(3), 5-74.

Bebbington

Brown

Frame

(2007). Accounting technologies and sustainability assessment models. Ecological Economics, 61, 224-236.

Bebbington

Brown

Frame

Thomson

(2007). Theorizing engagement: The potential of a critical dialogic approach. Accounting, Auditing & Accountability Journal, 20, 356-381.

Bekkers

(2013). E-government and innovation: The socio-political shaping of ICT as a source of innovation. In Osborne

Brown

(Eds.), Handbook of innovation in public services (pp. 253-268). Cheltenham, UK: Edward Elgar.

Bekkers

Homburg

(2005). E-government as an information ecology. In Bekkers

Homburg

(Eds.), The information ecology of e-government (pp. 1-19). Amsterdam, The Netherlands: IOS Press.

Berelson

(1952). Content analysis in communication research. New York, NY: Free Press.

10.

Bonsón

Ratkai

(2013). A set of metrics to assess stakeholder engagement and social legitimacy on a corporate Facebook page. Online Information Review, 37, 787-803.

11.

Bovaird

(2006). Developing new relationships with the “market” in the procurement of public services. Public Administration, 84, 81-102.

12.

Box

(1998). Citizen governance. Thousand Oaks, CA: Sage.

13.

Brabham

(2008). Crowdsourcing as a model for problem solving: An introduction and cases convergence. International Journal of Research into New Media Technologies, 14, 75-90.

14.

Brainard

Edlins

(2015). Top 10 U.S. municipal police departments and their social media usage. The American Review of Public Administration, 45, 728-745.

15.

Brown

(2009). Democracy sustainability and dialogic accounting technologies: Taking pluralism seriously. Critical Perspectives on Accounting, 20, 313-342.

16.

Brown

Dillard

(2013a). Agonizing over engagement: Social and environmental accounting and the “death of environmentalism” debates. Critical Perspectives on Accounting, 24, 1-18.

17.

Brown

Dillard

(2013b). Critical accounting and communicative action: On the limits of consensual deliberation. Critical Perspectives on Accounting, 24, 176-190.

18.

Bryman

Bell

(2015). Business research methods. New York, NY: Oxford University Press.

19.

Caron

Turcotte

M. B.

(2009). Path dependence and path creation: Framing the extra-financial information for the sustainable trajectory. Accounting, Auditing & Accountability Journal, 22, 272-297.

20.

Chua

A. Y. K.

Goh

D. H.

Ang

R. P.

(2012). Web 2.0 applications in government websites: Prevalence, use and correlations with perceived website quality. Online Information Review, 36, 175-195.

21.

Clarkson

M. B. E.

(1995). A stakeholder framework for analyzing and evaluating corporate social performance. Academy of Management Journal, 20, 92-118.

22.

Clarkson

P. M.

Overell

M. B.

Chapple

(2011). Environmental reporting and its relation to corporate environmental performance. Abacus, 47, 27-60.

23.

Cormode

Krishanmurthy

(2008). Key differences between Web 1.0 and Web 2.0. First Monday–Peer Reviewed Journal of the Internet, 13(6), 1-19.

24.

Coupland

(2007). Corporate social and environmental responsibility in web-based reports: Currency in the banking sector. Critical Perspectives on Accounting, 17, 865-881.

25.

Dahlberg

(2005). The Habermasian public sphere: Taking difference seriously? Theory and Society, 34, 111-136.

26.

Dawes

S. S.

(2008). The evolution and continuing challenges of e-governance. Public Administration Review, 68, 86-102.

27.

Deegan

(2002). The legitimising effect of social and environmental disclosures—A theoretical foundation. Accounting, Auditing & Accountability Journal, 15, 282-311.

28.

Deegan

(2006). Legitimacy theory. In Hogue

(Ed.), Methodological issues in accounting research: Theories and methods (pp. 161-182). London, England: Spiramus Press.

29.

Denhardt

J. V.

Denhardt

R. B.

(2007). The new public service: Serving, not steering. Armonk, NY: M.E. Sharpe.

30.

Denhardt

J. V.

Denhardt

R. B.

(2011). The new public service: Serving, not steering (3rd ed.). Armonk, NY: M.E. Sharpe.

31.

Denhardt

R. B.

Denhardt

J. V.

(2008). The new public service: Serving rather than steering. In Roberts

(Ed.), The age of direct citizen participation (pp. 63-77). Armonk, NY: M.E. Sharpe. (Original work published 2000)

32.

Dillard

Brown

(2012). Agonistic pluralism and imagining CSEAR into the future. Social and Environmental Accountability Journal, 32, 3-16.

33.

Dillard

Roslender

(2011). Taking pluralism seriously: Embedded moralities in management accounting and control systems. Critical Perspectives on Accounting, 22, 135-147.

34.

Dillard

Ruchala

(2005). The rules are game: From instrumental rationality to administrative evil. Accounting, Auditing & Accountability Journal, 18, 608-630.

35.

Etzioni

(1993). The spirit of community: The reinvention of American society. New York, NY: Simon & Schuster.

36.

Facebook quarterly earnings slides Q4 2015. (2016, January). Facebook, p. 5.

37.

Farazmand

(2012). Sound governance: Engaging citizens through collaborative organizations. Public Organization Review, 12, 223-241.

38.

Feeney

M. K.

Welch

E. W.

(2016). Technology-task coupling exploring social media use and managerial perceptions of e-government. The American Review of Public Administration, 46(2), 162-179. doi:10.1177/0275074014547413

39.

Fuchs

(2008). Internet and society: Social theory in the information age. New York, NY: Routledge.

40.

Gray

(2002). The social accounting project and accounting organizations and society: Privileging engagement, imaginings, new accountings and pragmatism over critique? Accounting, Organizations and Society, 27, 687-708.

41.

Greve

(2015). Ideas in public management reform for the 2010s: Digitalization, value creation and involvement. Public Organization Review, 15, 49-65.

42.

Habermas

(1984). The theory of communicative action. 1: Reason and the rationalization of society. Boston, MA: Beacon Press.

43.

Habermas

(1987). The theory of communicative action, volume 2: Lifeworld and system: A critique of functional reason. Boston, MA: Beacon Press.

44.

Habermas

(1989). The structural transformation of the public sphere. Cambridge, UK: Polity Press.

45.

Haveri

(2006). Complexity in local government change: Limits to rational reforming. Public Management Review, 8(10), 31-46.

46.

Holzer

Manoharan

(2008). Opportunities for civic engagement. In Yang

Bergrud

(Eds.), Civic engagement in a network society (p. 299). Charlotte, NC: Information Age.

47.

Kaplan

A. M.

Haenlein

(2009). Consumers, companies, and virtual social worlds: A qualitative analysis of second life. Advances in Consumer Research, 36, 873-874.

48.

Kaplan

A. M.

Haenlein

(2010). Users of the world, unite! The challenges and opportunities of social media. Business Horizons, 53, 59-68.

49.

Kaplan

A. M.

Haenlein

(2012). Social media: Back to the roots and back to the future. Journal of Systems and Information Technology, 14, 101-104.

50.

Kent

M. L.

Taylor

White

(2003). The relationship between Web site design and organizational responsiveness to stakeholders. Public Relations Review, 29, 66-77.

51.

King

Zanetti

(2005). Transformational public service: Portraits of theory in practice. New York, NY: M.E. Sharpe.

52.

King

C. S.

Feltey

K. M.

Susel

B. O.

(2008). The question of participation: Toward authentic public participation in public administration. In Roberts

(Ed.), The age of direct citizen participation (pp. 383-400). Armonk, NY: M.E. Sharpe. (Original work published 1998)

53.

Kraemer

King

(1986). Computing and public organizations. Public Administration Review, 46, 488-496.

54.

Landsbergen

Park

(2011, June 2-4). Managing network communications through social media. Paper presented to the Public Management Research Conference in Syracuse, New York, NY. Retrieved from http://www.maxwell.syr.edu/deans.aspx?id=77309416873

55.

Laughlin

R. C.

(2007). Critical reflections on research approaches accounting regulation and the regulation of accounting. British Accounting Review, 39, 271-289.

56.

Laughlin

R. C.

(1987). Accounting systems in organizational contexts: A case for critical theory. Accounting Organizations and Society, 12, 479-502.

57.

M. H.

Feeney

M. K.

(2014). Adoption of electronic technologies in local U.S. governments distinguishing between e-services and communication technologies. The American Review of Public Administration, 44, 75-91.

58.

Lovejoy

Waters

R. D.

Saxton

G. D.

(2012). Engaging stakeholders through Twitter: How non-profit organizations are getting more out of 140 characters or less. Public Relations Review, 38, 313-318.

59.

(2014). Diffusion and assimilation of government microblogging: Evidence from Chinese cities. Public Management Review, 16, 274-295.

60.

Manetti

(2011). The quality of stakeholder engagement in sustainability reporting: Empirical evidence and critical points. Corporate Social Responsibility and Environmental Management, 18, 110-122.

61.

Manetti

Bellucci

(2016). The use of social media for engaging stakeholders in sustainability reporting. Accounting, Auditing & Accountability Journal, 29(8).

62.

Meijer

A. J.

Thaens

(2010). Alignment 2.0: Strategic use of new internet technologies in government. Government Information Quarterly, 27, 113-121.

63.

Meijer

A. J.

Thaens

(2013). Social media strategies: Understanding the differences between North American police departments. Government Information Quarterly, 30, 343-350.

64.

Meijer

A. J.

Torenvlied

(2014). Social media and the new organization of government communications: An empirical analysis of Twitter usage by the Dutch police. The American Review of Public Administration, 46, 143-161. doi:10.1177/0275074014551381

65.

Mergel

(2010). Government 2.0 revisited: Social media strategies in the public sector. PA Times, 33, 7-10.

66.

Mergel

(2012). Social media in the public sector. San Francisco, CA: Jossey-Bass.

67.

Mergel

(2013). A framework for interpreting social media interactions in the public sector. Government Information Quarterly, 30, 327-334.

68.

Mitchell

R. K.

Agle

B. R.

Wood

D. J.

(1997). Toward a theory of stakeholder identification and salience: Defining the principle of who and what really counts. Academy of Management Review, 22, 853-886.

69.

Nabatchi

(2010). Addressing the citizenship and democratic deficits: Exploring the potential of deliberative democracy for public administration. American Review of Public Administration, 40, 376-399.

70.

Nalbandian

(2008). Facilitating community, enabling democracy: New roles for local government managers. In Roberts

(Ed.), The age of direct citizen participation (pp. 49-62). Armonk, NY: M.E. Sharpe. (Original work published 1999)

71.

Newman

Clarke

(2009). Publics, politics and power. London, England: Sage.

72.

O’Dwyer

(2005). The construction of a social account: A case study in an overseas aid agency. Accounting, Organizations and Society, 30, 279-296.

73.

Dahlberg

(2001). Computer-mediated communication and the public sphere: A critical analysis. Journal of Computer-Mediated Communication, 7(1), 1-16.

74.

Osborne

(Ed.). (2010). The new public governance? London, England: Routledge.

75.

Park

Reber

B. M.

(2008). Relationship building and the use of web sites: How Fortune 500 corporations use their Web sites to build relationships. Public Relations Review, 34, 409-411.

76.

Patten

Guidry

(2010). Market reactions to the first-time issuance of corporate sustainability reports: Evidence that quality matters. Sustainability Accounting, Management and Policy Journal, 1, 33-50.

77.

Pfeffer

Salancik

(1978). The external control of organizations: A resource dependence perspective. New York, NY: Harper & Row.

78.

Porter

M. E.

(2001, March). Strategy and the Internet. Harvard Business Review, 79(3), 62-79.

79.

Power

Laughlin

R. C.

(1996). Habermas law and accounting. Accounting, Organizations and Society, 21, 441-465.

80.

Ramanadhan

Mendez

S. R.

Rao

Viswanath

(2013). Social media use by community-based organizations conducting health promotion: A content analysis. BMC Public Health, 13(1), 1129.

81.

Roberts

N. C.

(2008). The age of direct citizen participation. Armonk, NY: M.E. Sharpe.

82.

Rosener

J. B.

(2008). Making bureaucrats responsive: A study of the impact of citizen participation and staff recommendations on regulatory decision making. In Roberts

(Ed.), The age of direct citizen participation (pp. 374-382). Armonk, NY: M.E. Sharpe. (Original work published 1978)

83.

Rybako

Seltzer

(2010). Dialogic communication in 140 characters or less: How Fortune 500 companies engage stakeholders using Twitter. Public Relations Review, 36, 336-341.

84.

Streck

J. M.

(1998). Pulling the plug on electronic town meetings: Participatory democracy and the reality of the usenet. In Toulouse

Luke

T. W.

(Eds.), Edshe politics of cyberspace: A new political science reader (pp. 18-47). New York, NY: Routledge.

85.

Suchman

(1995). Managing legitimacy: Strategic and institutional approaches. Academy of Management Review, 20, 517-610.

86.

Surowiecki

(2004). The wisdom of crowds. New York, NY: Little, Brown.

87.

Svendsen

(1998). The stakeholder strategy: Profiting from collaborative business relationships. San Francisco CA: Berrett-Koehler.

88.

Swift

Owen

D. L.

Humphrey

(2001). The management information systems dimensions of social and environmental accounting and accountability. London, England: CIMA Research.

89.

Tate

W. L.

Ellran

L. S.

Kirchoff

J. F.

(2010). Responsibility reports: A thematic analysis related to supply chain management. Journal of Supply Chain Management, 46(1), 19-44.

90.

Teicher

Alam

Van Gramberg

(2006). Managing trust and relationships in PPPs: Some Australian experiences. International Review of Administrative Sciences, 72, 85-100.

91.

Transit Cooperative Research Program. (2012). Uses of social media in public transportation. A Synthesis of Transit Practice. Washington, DC: National Academy of Sciences.

92.

Twitter Q4 2015 company metrics. (2016, January). Twitter, p. 1.

93.

Unerman

Bennett

(2004). Increased stakeholder dialogue and the internet: Towards greater corporate accountability or reinforcing capitalist hegemony? Accounting, Organizations and Society, 29, 685-707.

94.

Ventriss

(2008). Toward a public philosophy of public administration: A civic perspective of the public. In Roberts

(Ed.), The age of direct citizen participation (pp. 38-48). Armonk, NY: M.E. Sharpe. (Original work published 1989)

95.

Villa

D. R.

(1992). Postmodernism and the Public Sphere. American Political Science Review, 86, 712-721.

96.

Voss

Z. G.

Voss

G. B.

Moorman

(2005). An empirical examination of the complex relationships between entrepreneurial orientation and stakeholder support. European Journal of Marketing, 39, 1132-1150.

97.

Waddock

S. A.

(2002). Leading corporate citizens: Vision, values, value added. Boston, MA: McGraw-Hill.

98.

Waters

R. D.

Jamal

J. Y.

(2011). Tweet, tweet, tweet: A content analysis of non-profit organizations Twitter updates. Public Relations Review, 37, 321-324.

99.

Windsor

(2002). Stakeholder responsibilities: Lessons for managers. In Andriof

Waddock

Husted

Rahman

(Eds.), Unfonding stakeholder thinking: Theory, responsibility and engagement (pp. 137-154). Sheffield, UK: Greenleaf Publishing.

100.

Xifra

Grau

(2010). Nanoblogging PR: The discourse on public relations in Twitter. Public Relations Review, 36, 171-174.