PUBLIC INFORMATION OFFICERS AND SOCIAL MEDIA: NEW MEDIA, NEW METHODS? By Kourtnie R. Rodgers A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Criminal Justice – Doctor of Philosophy 2024 ABSTRACT Police public information officers (PIOs) serve as spokespersons for their departments and are often considered gatekeepers of information to the public. PIOs work with media to ensure information is put out in a timely manner and to manage department image. The media landscape has drastically shifted in the past decade, particularly with the introduction and popularity of social media. The role of the PIO has also continued to become more professionalized as departments realized the importance of the position. Prior to this study taking place, two major events took place: the COVID-19 pandemic and the murder of George Floyd. These events cannot be ignored and may have impacted the role of PIOs as well as their use of media. With these advances and context in mind, this dissertation sought to fill the dearth of information about the current role of the police PIO in the U.S., to examine how organizational and institutional pressures may impact this role, and to understand PIOs’ responses to a few high-profile cases. To accomplish this goal, a national survey of PIOs was conducted (N=144) and linked with collection of public social media data and organizational characteristics. Analysis of the survey results revealed that PIOs now use and prefer social media for many job tasks, but still also use more traditional media. Social media also seems to have removed the barrier of traditional media gatekeepers, allowing for PIOs to craft their messages, responses, and department image more directly. This study also found individual and organizational factors that influence PIOs social media outputs include PIOs perception of their competence with social media, enjoyment of social media, view of importance of social media, diversity of the area served, and size of the agency. Finally, the major events of 2020 were found to bring stress to the role of the PIO. COVID-19 moved even more of the PIOs’ role online and required a shift in engagement strategy. In a shift back, the murder of George Floyd required PIOs to connect with their community again, sometimes offline. This dissertation serves as an update to our understanding of the role of the PIO and a beginning to our understanding of how PIOs responded to drastic shifts in the media environment and how organizational theory components may impact PIOs’ use of social media. Finally, this dissertation was uniquely situated to capture PIOs’ responses to the high-profile events of 2020, which highlighted how major events impact the role of the PIO and how important context is to their work. Copyright by KOURTNIE R. RODGERS 2024 In loving memory of my Grandpa Jerry and Grandma Myrna. v ACKNOWLEDGEMENTS I have so many thank yous to give. I want to start by thanking my family. We have all had a rough few years getting here and I hope this accomplishment is a bit of a bright spot and start to a new chapter. Thank you to my husband Matt who was constantly a source of encouragement and love when I doubted myself (which was often). Thank you to my son Adler who brightens all my days and who motivated me to finish to spend more time with him. Thank you to my parents, Anna and Brian, who always encouraged me to be my best in whatever I chose to do and who taught me to be resilient and brave. Thank you to my aunt Tobi who was also there with support and feedback on my presentation. Thank you to all my extended family who provide me with joy and prayers from afar. You are all my inspiration and my strength; I hope I can be the same for you. I also need to thank my committee who brought so much knowledge and support to this project. Thank you to my chair Dr. Steven Chermak for your true patience, all the time you spent on this project, and for encouraging me throughout the whole process. Many days I did not think anyone beyond me cared about this study and you reminded me that it matters. I appreciate and thank my committee members, Drs. Thomas Holt, Kjerstin Thorson, and Jeremy Wilson. Thank you for inspiring me and this work. Thank you all for your invaluable input that made this dissertation better and for the time you spent reviewing it. Finally, thank you to all my friends who supported me at Boise State, MSU, and in my job at ISP. You have all made my graduate school years fun, challenging, and successful. Thank you all for joining me on this long journey and encouraging me every step of the way. The lively discussion, coffee, and walks through campus kept me going. vi TABLE OF CONTENTS CHAPTER 1: INTRODUCTION ....................................................................................................1 CHAPTER 2: RELEVENT LITERATURE ..................................................................................23 CHAPTER 3: METHODOLOGY .................................................................................................56 CHAPTER 4: SURVEY RESULTS ..............................................................................................69 CHAPTER 5: ORGANIZATIONAL THEORY FINDINGS........................................................85 CHAPTER 6: HIGH PROFILE CASES .....................................................................................106 CHAPTER 7: CONCLUSION ....................................................................................................114 REFERENCES ............................................................................................................................127 APPENDIX ..................................................................................................................................132 vii CHAPTER 1: INTRODUCTION This project explores the role and work of police public information officers (PIO). Particularly, this project updates the existing understanding of PIOs, examines PIOs’ management of social media, assesses factors that may influence PIOs use of social media, and investigates PIO management of extraordinary events. How has the role of police PIOs changed with the drastic change in the media landscape? Have departments become more reliant on social media as a form of quick communication with the public? Do police departments and PIOs still focus on traditional media outlets to push out information? Has social media changed police department relationships with other media? Are police department PIOs able to connect better with the public they serve through social media? What department characteristics impact PIOs use of social media? How do PIOs respond to extraordinary events? These questions are explored by examining PIOs' survey responses and their activities on department social media pages. While there is research observing reactions to other high-profile cases, the events of 2020 were unprecedented. This project fills substantial research gaps by exploring the current methods, strategies, resources, and forms of communication used by police PIOs, how social media has changed the PIO role, and how department and individual-level variables impact PIO perceptions and use of social media. Further, this project examines the impact on PIOs of two unprecedented 2020 events: the COVID-19 pandemic, and the protests against police violence after the death of George Floyd. This study combines survey data with social media content from a national sample of police department PIOs. Two data-gathering techniques were utilized: an online, self- administered survey of 582 police PIOs serving populations over 50,000 or employing at least one hundred sworn officers in the U.S. and a collection of public social media data from 1 responding departments. The sampling frame was selected from the 2016 Law Enforcement Agency Roster(LEAR) which contains one of the most complete lists of agencies in the U.S. (DOJ, 2016). The survey measures the role of the PIO with items adapted from Chermak and Weiss (2006) along with newer items that were developed specifically to capture: a) methods used to market police departments on social and traditional media, b) the personnel responsible for this work, c) perceived benefits and drawbacks of using social media, d) training and knowledge of various social media platforms, e) organizational characteristics that may impact social media use, f) public relation responses to the COVID-19 pandemic, and g) public relation responses following the death of George Floyd. The perspective of PIOs and their relationship with and understanding of social media is an area that has yet to be explored in significant detail. This survey provides invaluable insight into the perceptions of the creators of police social media content-an important, yet little-studied area. These data were then attached to a collection of the agencies’ publicly available Facebook data to allow for further understanding and examination of outputs concerning the PIO's reported inputs. The social media data collected for each department includes account data involving the number of followers, posts, and reactions over four years, from January 1, 2019, through January 1, 2023. This period was intentionally created to capture a baseline before major events, COVID-19, before the death of George Floyd, after the death of George Floyd, and data during and after the time of survey data collection. Data from the 2015 IACP survey indicates that the platform most used by police departments was Facebook (94.2%), increasing the likelihood that the sample would have activity on this platform. Integrating Facebook data with the survey data from PIOs has not been done previously and provides many opportunities to explore research questions that have yet to be explored. 2 Data about the account such as the number of followers, number of posts in a specified time range, and reactions collected provides context and understanding of the audience on the platform. This study utilizes adapted social media metrics proposed by Bonsón et al. (2017) and Haro-de-Rosario (2018) to measure department activity and dimensions of citizen engagement. While these are not perfect measures of reach, as many more people may see a post and not be followers or react to the post, these measures do give a good sense of the audience that has the potential to see posts regularly and those who choose to actively engage with the department. The combination of the survey and social media data methods provides crucial insight into the management and development of police social media communications through the PIOs, a major piece that is missing in research up to this point. The survey data and social media data were collected concurrently to allow for responses to be attached to the outputs observed on social media. Social media data was also collected before the survey collection period to provide baseline data that may not be related to the survey. This chapter focuses on three issues: First, the importance and relevance of this study will be reviewed, along with a discussion of the environment this study was being conducted in. Second, the research questions for the study will be presented. Finally, an outline of the remaining chapters will be presented. Importance Police and media Decades of research have been devoted to the understanding of media and its impact on fear of crime and attitudes towards the criminal justice system. Crime has been a popular subject for media and remains popular among various audiences. Crime news specifically has experienced a general increase in the percentage of news coverage since 2010 (Tyndall Report, 3 2019). Moving beyond news media, other media has also been inundated with crime content. Crime-related T.V. dramas, documentaries, and podcasts have experienced success as well with the top three most watched T.V. shows of the 2019-2020 season being the crime procedural dramas NCIS, FBI, and Blue Bloods (Clark, 2020). These three shows alone had 39.81 million viewers. While crime media is certainly prolific, less research has been devoted to the examination of media impacts on perceptions of police. Media, in general, are known to have an impact on perceptions of police (Callanan & Rosenberger, 2011; Chermak et al., 2006; Dowler & Zawilski, 2007). Police organizations have historically had to manage their relationship with media organizations as a way to control their information and image. Further, police image is crucial to maintaining organizational legitimacy in the eyes of the public (Chermak & Weiss, 2006). Traditionally due to prevalence, media have provided police departments with access to a wide audience to distribute multiple types of information such as crime information or information about public safety (Chermak & Weiss, 2006; Ericson et al., 1989). News media and police have historically held generally positive relationships as each benefit from the other. News media gain the information they need, and police have rapid access to a wide audience if needed. This can become particularly important in times of emergency when police need to make mass notifications in specific areas and when news media need reliable sources for up-to-date information. As media have evolved, it is evident that departments have altered their communication strategies to fit with new media and understand the importance of media to their legitimacy. New media have introduced diverse ways for police to communicate quickly with the population they serve and beyond. Technology has made it possible for messages to be produced, distributed, and consumed 24/7. When only pushing information out through news media, police departments’ 4 information is subject to the editing, production, and time structures of the news media. News organizations are guided by what is of interest to consumers and thus pick and choose what they believe is valuable and will sell (Chermak, 1994). Police are now, through social media, better able to produce and distribute information without such constraints, no longer being filtered through news media organizations. Police organizations are also able to receive and evaluate direct feedback and reactions to the information they disclose. Valuable resources including training, time, and money, have been put toward maintaining the police image through various types of strategies for these new media. While there is a decent amount of research examining police communication with the public, there is little examining the impact of the current media structure and how this may have changed the work of those most directly involved in managing police image. One strategy used by the police to improve their external communication is the hiring of PIOs who act as spokespersons for the department. Previous research has found that PIOs and traditional media groups report elevated levels of benefits from positive relationships with each other (Chermak & Weiss, 2006). While the PIO acts as a gatekeeper of information to the public, traditional media such as news and radio also are useful to spread information to a wide audience (Surette & Richard, 1995). More recently, this relationship may have become less important as new media have become popular. As the public has found innovative ways of gathering and obtaining information, PIOs may have adjusted their tactics to maintain legitimacy with a wide audience. Citizens utilize added information gathering strategies and may now be more likely to gather their information from open electronic sources online. A 2020 survey found 26% of U.S. adults claiming to get their news from the popular social media platform, YouTube (Stocking et al., 2020). Further, the public may also be losing trust in news media and question the quality of 5 information provided in the news and the intentions behind journalists’ work (Gottfried et al., 2020). With such a drastic shift in the media landscape and information-seeking behaviors of the public, and the potential that these shifts have impacted how police engage the public, there is little understanding of the current strategies and methods used by police PIOs to communicate with the public. Updating understanding of PIOs Media has drastically shifted in the past decade in many ways that should have impacted the job of the PIO. Media have become much less centralized and have expanded to include multiple types of producers and contributors. Importantly, there has been a decline in local news media, creating “news deserts” in which the local media is absent, which may lead to information needs of a community not being met (Ferrier, Sinha, & Outrich 2016). From 2004- 2019 the number of newspapers in the United States decreased substantially (Abernathy, 2020). From 2018-2020, 300 newspapers closed, 6,000 journalists lost their jobs, and print newspaper circulation declined by 5 million (Abernathy, 2020). Voids of local information are created as local media businesses struggle through fiscal crisis and lack the necessary sources to support themselves through the introduction of the internet and new media mega-chains of large publicly traded newspaper companies (Abernathy, 2020). The shift has led to a model of news making that is focused on profitability by large companies who have no ties to the local community (Abernathy, 2020). Social media has certainly impacted the news business as well as most aspects of public life (Abernathy, 2020). The delivery of news and information has changed dramatically, with new contributors stepping in to create content at lower costs through online platforms. With this massive change in how the media functions, particularly at the local level, there is a need to 6 update our understanding of the job of the PIO. Previously, the PIO worked closely with local media and held close relations with local news media producers (Chermak & Weiss, 2006). As news organizations fade away or have much fewer reporters dedicated to creating specific types of news, and local news has declined in importance, how have PIOs managed to continue to communicate with the public and manage the image of the police? How are police organizations strategically using PIOs to navigate the new media marketplace? Critical role of the PIO PIOs play an integral part in police departments’ communication with the public. The PIO has been described as a gatekeeper in the “flow of crime and justice information” to the public (Surette & Richard, 1995). They are the main output of information from the department and have been institutionalized in policing for decades now. The PIO is a specialist in the organization and the face of the department to the community who dictates the communication strategy of the department in consultation with command staff (Chermak & Weiss, 2006; Skolnick & McCoy, 1985). PIOs ensure consistency in messaging. They are also responsible for ensuring that crucial public information is put out in a timely and accurate fashion. Image management has potentially become easier as PIOs are allowed to curate their image online and more carefully craft the messages that the public eventually receives, without having to manage the news media filtering process. However, news organizations remain a force within these communities and continue to cover criminal justice issues in some detail. Clearly, the role of the PIO is important to police departments who must maintain a professional image and legitimacy with the community. As recognition of the importance of this position and its function in police departments has grown, there has also been a rise in national and statewide professional associations, further 7 signaling the importance of this position. For example, associations such as the National Information Officers Association (NIOA) provide PIOs with training, development opportunities, networking opportunities, and annual conferences. Further, the International Association of Chiefs of Police (IACP) has established a section of the organization devoted to PIOs and hosts a conference every year bringing together PIOs to exchange ideas and provide training opportunities. The PIO position within police departments is becoming more specialized and professional as the importance of the position is realized. The topics at this conference also highlight the issues that are particularly pressing for PIOs: “Telling your story for successful social recruitment”, “Public Information “Reporter”: How to integrate storytelling into public messaging”, “Social Media Etiquette and Ethics”, “When Social Media Sucks,” and multiple presentations on messaging during high profile cases. While the role of the PIO has grown and become more professionalized, little research has examined this shift. Prior research on the role of PIOs found that the majority of their time (44%) was spent providing information about crime incidents (Chermak & Weiss, 2006). The next largest percentage of their time (15%) was spent providing information about police initiatives such as community policing, and the rest was spent providing information in other areas including police misconduct, award ceremonies, and organizational practices (Chermak & Weiss, 2006). As the media landscape has shifted it is likely that the priorities and responsibilities of PIOs have also shifted. One might expect that there is more time spent on what the department deems important, rather than responding to requests for information. For example, if the department is focused on community policing initiatives, public participation is necessary. As the PIO acts as the public face of the department, it is likely that they may be the only contact with the police department many individuals have. The PIO then may want to make 8 themselves accessible and even more visible to the public to manage department appearance well. As PIOs are utilizing new mediums, they may also be spending more time developing and producing content that is more consumable than prior press releases. This may include having more pictures or video content to curate as new media such as social media introduces the ability to post content of multiple types in a single post. Where news media previously took a press release and curated the content, now, a PIO may curate their own content, involving a whole new set of skills than was previously required for the job. The PIO position has become more important as law enforcement in general grows to involve a more open system of communication (Motschall & Cao, 2002). While the PIO previously had to compete with the priorities of media producers, new media provides potentially more freedom for the types and amount of content PIOs can create, thus potentially impacting their strategy and resource distribution. A further understanding of the PIO position in current times is important to the understanding of police communication and community relationships as a whole. Lack of citizen support may be detrimental to police in many ways as police rely on citizens for cooperation in crime initiatives but also resources and budget support. An increase in understanding will provide information to further improve the role of the PIO, and potentially improve community relationships, by highlighting areas where PIOs may need further education or training and budget needs or areas that can be cut. PIO management of social media Role of social media Social media is defined as a group of internet technologies that enable people to interact with others and allow users to “create, edit, evaluate, and/or link to content” to other creators (Kaplan & Haenlein, 2010; van Der Graaf, 2015). Social media are easily accessible, and 9 available on mobile platforms so that they can be accessed almost anywhere at almost any time. One of the most prolific social media companies, Meta Platforms, Inc., began in February of 2004 as Facebook, Inc. and has since grown to be a multi-billion-dollar company employing over 67,000 employees with offices in cities worldwide (as of December 2023) (Meta, 2024). Meta also has had a steady increase in users, with the latest count of Facebook daily active users at 2.11 billion on average for December 2023, a 6% increase compared to 2022 figures (Meta, 2024). X, formerly Twitter, another popular social media platform, reported a 34% year-over- year growth in their daily active users in Q2 of 2020, growing to 186 million, thirty-six million of those being in the U.S. (Twitter, 2020). Since Twitter was purchased by Elon Musk in 2022, the company is now privately held and no longer releases quarterly earnings but the last quarterly statement in Q2 of 2022 indicated a loss in revenue but an increase in daily active users up to 237.8 million, with 41.5 million in the U.S. (Twitter, Inc. 2022).Social media is truly redefining communication at all levels of society, both inside and outside of organizations and the government. Police are no exception to the growth in social media use. The International Association of Chiefs of Police (IACP) annual survey on social media found at least 81.1% of responding police agencies using social media since 2010, rising to 96.4% of respondents in 2015 (IACP, 2010; IACP, 2015). Unfortunately, this survey has not been repeated to measure how many agencies report using social media now. Users of social media play an active role as consumers, producers, and circulators of information, unlike traditional media where users play the role of consumer. The complexities of social media create a space that is shaped and continually molded by multiple individual users as well as companies who own the platforms. There is potential to provide one-to-one, one-to-many, and many-to-many interactions within these digital spaces 10 (van der Graaf, 2015). Social media provides a new venue for communication and potentially a more direct channel between police organizations and the community. The affordances of social media make it a useful tool for local government users, including the police, who wish to interact with citizens, especially as the public increases in demands for information delivery, transparency, and representation. However, research up to this point on local government use of social media has observed that these organizations largely use social media to self-promote and as political marketing rather than increase transparency and participation (Medaglia & Zheng, 2017). The slow uptake of these features has been attributed to a lack of understanding of social media and policy, especially as some of this research took place as government agencies were just beginning their use of social media. Thus, it is expected that more interactive use of social media may increase as understanding of the platforms and their potential increases, but transferring policy and practices to a digital space can be complicated for certain groups to implement. Police may utilize social media in many ways that link them directly to the citizens they serve such as warning their community about threats, encouraging citizens to report problems, and even involving citizens in investigations (Dekker et al., 2020). This is further supported by the finding that the top three agency uses of social media reported by police chiefs in the 2016 Law Enforcement Use of Social Media Survey were “Notifying public of public safety concerns”, “Community outreach and engagement”, and “Public Relations” (Kim et al., 2017). The flow of information from the PIO has generally been found to be one-way and reactive but there has also been some support for the emergence of two-way interaction, with PIOs serving to assess public opinion to inform internal actions through examination of social media (Motschall & Cao, 2002). 11 Social media provides a free, and familiar platform for citizens and police to interact. It may be that two-way forms of communication have increased as more police departments have attempted to solicit public opinion through social media, yet this remains unexplored. While in- person encounters with police are clearly important, these online communication relationships and encounters with local police are understudied. There is potential for multiple forms of information flow through online channels, particularly social media, and police organizations may choose what best fits their perceived needs or goals. The importance of this line of research is highlighted in the final report of The Presidents’ Task Force in 21st Century Policing, a group that was assembled by President Obama partially in response to unrest in Ferguson, Missouri following the shooting of Michael Brown by a police officer. Pillar 3 of this report is titled “Technology & Social Media”, and the report encourages the use of social media by departments as a tool for engagement and education of their local communities and to increase community trust and access (President’s Task Force, 2015). The report also underscores the fact that social media use may be a great tool for engagement and information sharing, but it must be done professionally, thus highlighting the importance of having someone like a PIO in control of such media. The professionalization and management of social media requires that there be policy and a person who is trained in public relations in charge of media. While some prior literature has analyzed the role of a PIO in police organizations (Chermak & Weiss, 2006; Motschall & Cao, 2002), little research has examined how or if this role has changed since the introduction of social media. Social media channels have altered the communication landscape and introduced new opportunities and challenges to the position of the PIO. PIOs must learn how to be an even more direct source of information for the public, balancing their needs with the public’s wants for information. As such, many questions remain as 12 to what departments are doing to maintain legitimacy with the public through the PIO role, particularly as unrest and distrust in police in the U.S. may be increasing drastically. The PIO is expected to maximize the positive and minimize the negative images about their organizations (Chermak & Weiss, 2005). With social media, this role has potentially become easier in that they can curate their image on their pages, but it also has brought new difficulties with an endless flow of information and images from multiple sources that may be harder to manage, particularly in times of response to high profile cases or unrest. Police organizations must be able to effectively communicate and respond in various situations including scandals, changes in crime rates, prolonged investigations of high-profile crimes, and increases in public criticism and blame (Chermak & Weiss, 2006). Response to such cases and crises is critical. PIOs, media, and Organizational theory Many things may influence how a PIO utilizes social media in their role. While studies have established that PIOs are using social media, and even examined what types of things are being posted to social media, little has moved on to understand the motivation behind what is posted and why. Up to this point, there has been little to no organizational theory testing of PIOs and their use of media, especially social media. There are, however, some studies that examine these variables in related topics that can be used as a base for our understanding of PIOs. Organizational characteristics and their relation to use of social media by government organizations in general have been examined, but even that line of research remains limited to a few studies (Medaglia & Shend, 2017). Police media strategy in general also has few studies that indicate that there may be internal factors determining social media strategy as well as contingencies such as agency size (Meijer & Thaens, 2013; Kim et al., 2017). These studies can provide background, yet testing of other factors remains limited. The ability of an organization to 13 adapt to the environment it operates in and respond to community needs impacts the effectiveness of the organization (Donaldson, 2001). To fill in these gaps in research, this study also aims to examine the organizational and institutional pressures that may impact how a PIO uses social media specifically. Applying organizational theory to the PIOs use of social media is an important step to further the research that has been done up to this point. PIOs and high-profile cases Social media is a valuable tool in police communication in the current era, particularly regarding police response to major events. Major events or celebrated cases are especially important to the criminal justice system in the U.S. as the system’s response to these cases goes well beyond the individuals involved. Sam Walker proposed, in his wedding cake model of criminal justice, that celebrated cases take up the majority of media attention while they truly represent a minority of cases (Walker, 2015). Police specifically tend to be the most visible of all criminal justice institutions to the public, through both primary interactions and secondarily through media (Chermak & Weiss, 2005; Goldsmith, 2010). Consuming news media has been found to impact citizen attitudes about the guilt of officers in a celebrated case, and the publicity of a police misconduct trial was observed to intensify concerns and negative perceptions among black individuals specifically (Chermak et al., 2006). It seems that the regular crime news may not necessarily impact perceptions of police, but celebrated cases are produced differently and the effects on audiences are noticeable (Walker, 2015). These effects may be different when the media consumed is not produced through centralized, established media creators such as news media, but through citizens themselves. Through advances in technology, the secondary visibility of policing has grown as citizens have increased access to cameras, video, and the internet (Goldsmith, 2010). One of the 14 first notable uses of technology by a citizen to record police actions that spread far beyond the individuals involved was the recording of the Rodney King incident in 1991 in Los Angeles. This recording and subsequent distribution demonstrated the potential of technology use by citizens and the impact it can have when exposing negative policing practices (Goldsmith, 2010). This example also shows how cases no longer must rely on a catalyst of being picked up by a major news outlet but are typically spread through multiple sources with different content filters. The use of technology by police, rather than the public, in response to major events has also been observed. For example, social media use in response to emergency events has been examined in the Boston Police Department’s use of Twitter after the Boston Marathon Bombing. A study found two-way communication occurring between the department and the public (Davis et al., 2014). This communication improved the response, allowed for a quick reaction to misinformation, and provided a space to get information about the suspects (Davis et al., 2014). The positive interactions between the police and public in an online space highlights the breadth of uses that social media may provide. Police may be able to gauge public sentiment through social media, as drastic differences have been observed through examination of public Twitter posts about police following crises such as Freddie Gray’s death and subsequent protests in 2015 (Oglesby-Neal et al., 2019). Sentiment analysis of social media, especially Twitter, has been completed after such events, yet little is known about how closely police agencies monitor public sentiment online themselves. While prior studies have been conducted regarding high-profile events, public sentiment has again drastically shifted following the death of George Floyd in May 2020 which was filmed and widely distributed online. Online police response to the incident and protests has not been examined, yet it seems clear that there is a shift in public sentiment in the U.S. This may affect 15 this study in altering how PIOs utilized social media because they were reacting to a major event. It is important to recognize the environment this study was implemented and understand that there are events that may have altered the use of social media by both the police and the public speaking about the police online. Another major event that might demonstrate how police PIOs react to external stimuli is the COVID-19 pandemic. This public health crisis created an environment in which individuals needed information and police were expected to take on new tasks such as enforcing social distancing. Police also needed to implement their own measures to protect officers. In the July 25th edition of a weekly email sent to members of the Police Executive Research Forum (PERF), the executive director Chuck Wexler stated that “the last four months have been exhausting for police chiefs and officers” and chiefs are feeling “challenged or even bewildered.” The death of George Floyd and subsequent protests occurred during the pandemic creating an environment with “mixed and overlapping impacts” for police (Wexler, 2020). These events have impacted policing in the U.S., and thus likely impacted their communication strategies, and are part of the approach to managing these crisis events. With recent events, there is potential to examine how police are responding to a variety of events that have national impacts. Recent literature, examining public attitudes following high- profile events like Ferguson, revealed that police departments consider local media’s impact as more important, and potentially more harmful, than national media (Matusiak, 2019; Nix et al., 2020). The impact of the events of 2020 seems to spread far beyond local impact. This study attempts to reassess these findings in the current environment, which is the aftermath of another high-profile event, the death of George Floyd. The legitimacy of an organization can be drastically altered by what is presented in the media and police departments must manage this 16 through communication with local audiences. This communication must come from the local police and be directed at the citizens they serve and interact with on a day-to-day basis. Social media provides a platform to push out legitimate information quickly and to a specific or broad audience. Examination of the questions presented in this dissertation becomes particularly important as recent events provide further opportunities to observe PIOs' reactions to unprecedented events. Two major events occurred in 2020 that challenged police in general and may lead to interesting insights in how police PIOs manage department image and information through social media. The two events, the COVID-19 global pandemic and the death of George Floyd and subsequent protests, are significant enough that all departments in the study were likely impacted and required response in some way. The influence of these events on the U.S. cannot be understated and should highlight barriers and strengths that the PIOs experience in the use of social media. There could be multiple types of responses to these events that may be observed. For example, police could either push hard on image management or they could react by pulling away from social media if their efforts seem in vain. Studying the PIO role at this time provides a prime opportunity to understand the full scope of the job, particularly when under stress and met with unprecedented experiences. Research Questions In order to address the large gaps that are currently in the literature, this study examines police communication online, how efforts by police organizations to manage their image have been impacted by the increases in use of social media, and how organizational characteristics may impact this use. Similar issues are explored through the examination of reactions to the two events outlined about, COVID-19 and the death of George Floyd. The study thus accomplishes 17 three important objectives. First, this study documents the current work of PIOs within police organizations, updating results from a study that was completed over twenty years ago (Chermak & Weiss, 1998; 2006). Since this time, vast changes have occurred in the media landscape, particularly in the rise in popularity of social media and the amount of access people have to these media. There is potential that the relationship between police and media has changed. PIOs are the primary way through which police departments manage their image. A deeper understanding of the current role of the PIO may help inform departments for hiring, training, and resource allocation. Updating Understanding of PIOs RQ1: Who are the personnel responsible for marketing police? What type of training do the personnel responsible receive? RQ2: What are the current communication methods being used by police PIOs in the U.S.? What strategies are being used by PIOs to market police in the media? RQ3: What are the relationships between media and PIOs? What are the PIOs perceptions of portrayal of police by different media? Second, this study explores further PIOs management of social media specifically and what organizational variables might impact their use. Multiple researchers have pointed to a massive gap in the literature, the lack of research examining, interviewing, or otherwise observing the creators of police social media content, usually the PIO (Dekker et al., 2020; Hu & Lovrich, 2020). The majority of police social media research up to this point focuses solely on the public content being created rather than considering the person/people behind the content and the motivation. The study also explores variations in social media activity and engagement by law enforcement organizations in the U.S. A growing area of communication research is an 18 examination of how government agencies use social media. According to Medaglia and Zheng (2017), this research area focuses primarily on the impacts of how organizational-level variables impact social media usage. The variables that have been observed most often include size, media strategy, organizational policies, IT capacity, and human resources capacity (Medaglia & Zheng, 2017). Much of this research concludes that agency size influences the adoption and use of social media (Bonsón et al., 2017; Guillamón et al., 2016; Lev-On & Steinfeld, 2015; Lidén & Larsson, 2016; Ma, 2013; Srivastava, 2016). A clear gap to fill, as identified by Medaglia and Zheng (2017), involves investigating the relationship between organizational level variables, how an agency manages its media and social media efforts, and how the individual or individuals responsible for these efforts engage the media environment. This study was designed to address these gaps in police research specifically and adds nuanced understanding by examining how both organizational and individual level variables impact social media outcomes. The following research questions are addressed: PIO management of social media RQ4: How do PIOs utilize social media and how is this use similar and different from traditional media? RQ5: What strategies are being used to communicate with the public, do they vary based on format (social media vs. traditional media)? RQ6: What changes has social media brought to police communication, specifically, what are the benefits and drawbacks? RQ7: What department resources (training, funding, time) are being put towards social media and how do PIOs collaborate with other units in the organization to share or collect information via social media? 19 RQ8: Do organizational variables (e.g., size of the police organization, budget, extant policy, leadership; technological support/capacity) impact the PIOs perceptions of the value of social media? Do these variables impact the activity and engagement of social media use by an organization? RQ9: Do individual-level PIO variables (understanding of the platform, training and experience with social media, motivation) impact activity and engagement of social media use? RQ10: Are organizational or individual level variables better predictors activity and engagement of social media use? Third, this study examines, retrospectively, PIO responses to two high profile events occurring in 2020, the COVID-19 pandemic and the death of George Floyd by asking PIOs to consider how their job changed as well as examining social media use over time. These events are unique and bring the opportunity to further understand how police respond to such events. The study provides the opportunity to specifically examine social media usage and public engagement over time. While there has certainly been research on other high-profile events, each of these are completely unprecedented in their scope. Every department in the study had exposure that allowed them to respond to questions surrounding these events. PIOs and high-profile cases RQ11: How has COVID-19 impacted PIO strategy and use of social media, if at all? RQ12: How has the death of George Floyd impacted PIO strategy and use of social media, if at all? 20 RQ13: How did each of these events impact social media engagement overtime, if at all? Outline of Chapters The second chapter of this dissertation will further cover the relevant literature. First, elements of organizational theory including contingency, institutional, and isomorphism will be presented as they relate to police organizations. These concepts provide a solid framework to begin to understand the organizations PIOs are situated in. The history of research examining police relationships with media will be discussed, including the role of social media in police communication. Next, the role of the Public Information Officer in police organizations will be reviewed and what understanding needs further investigation. Finally, PIO responses to high profile cases will be highlighted with the focus on two major events of 2020, COVID-19, and the death of George Floyd. The COVID-19 pandemic introduces complexity into police communication as people look to government agencies to respond. While this event is unique, research on police and government health communication will be discussed. Police legitimacy and the police use of media during crises such as the murder of George Floyd and subsequent protests is also somewhat unprecedented, but research on how police use media during similar events, such as after the shooting of Michael Brown by police in Ferguson, Missouri will be discussed. The third chapter will review the methodology of the dissertation. This chapter will discuss, in-depth, the sampling procedure, survey development, and the instruments used. It will also discuss CrowdTangle, the platform used to collect public Facebook data. An explanation of how the survey data and social media data were linked will be provided. Finally, this chapter will outline the analytic strategy of the study and discuss the variables being used including 21 measurements of department activity, citizen engagement (popularity, virality, and commitment), and characteristics of the PIO and police organizations. The fourth, fifth, and sixth chapters are dedicated to the presentation of results. Chapter four presents an update on the state of the PIO position as a whole and will address research questions 1 through 6. Chapter five provides results in relation to the organizational and institutional variables examined and will address research questions 7 through 10. Chapter six will discuss the impact of 2020 events including the murder of George Floyd and the COVID-19 pandemic and will address research questions 11 through 13. Chapter 7 will include the discussion, limitations, and conclusions of the study and offer suggested future lines of research. 22 CHAPTER 2: RELEVANT LITERATURE Introduction This chapter is organized into several sections. The chapter begins with a review of organizational theory elements including contingency, institutional, and isomorphism, then moves onto a discussion of police and media research. It is important to begin with theory and media relations research as there is a solid base supporting the importance of organizational structure and the interactions between police agencies and media. The second main section of the chapter will review research on Public Information Officers (PIO), their importance to police departments, and the importance of updating our understanding of this position. Next, the limited police and social media research and findings will be discussed as well as research on organizational characteristics and use of social media. This section will conclude by further focusing on PIOs and social media specifically. The next section of the chapter will provide an overview of the literature on media and police reaction to high-profile cases and will briefly discuss the two major cases occurring during the timeframe of the study, COVID-19, and the death of George Floyd. Finally, hypotheses will be presented in the conclusion of the chapter. Organizational Theory While the role of the PIO is crucial, it is also important to recognize that the PIO acts within a larger organization and it is important to consider how organizational theories are relevant for understanding PIOs' actions and interactions with media and the public. There are three key theoretical areas that this study will consider including contingency, institutional, and isomorphism. 23 Contingency Structural contingency theory (SCT) has been immensely popular among policing scholars seeking to understand how police organizations respond and adapt to their environment. While contingency theory became popular in the 1960s and 1970s, scholars continue to use this theory in slightly different ways (Rainey, 2003). According to Donaldson (1995), organizations seek to increase and maintain their performance and to achieve this goal, organizations must respond to internal and external factors or contingencies. Environment, size, and strategy all impact organizational structure and if these contingencies change, organizations must respond (Donaldson, 2001). In policing, many scholars have examined the influence of typical contingency variables such as agency size, population served (size, growth, makeup), and technology on police innovation. It is important to note that contingency theory is not stating that organizations only respond structurally, but also in their behavior. Organizational Institutional theory Institutional theory posits that organizations focus on improving legitimacy by adopting structures and strategies that are supported by stakeholders and key figures. Meyer and Rowan (1977) argue that organizational structures reflect the myths of their institutional environment. Organizations seek to maintain this legitimacy and do not focus on organizational effectiveness but rather on survival in their environment through more symbolic means (Meyer & Rowan, 1977). In policing, goals and strategies may be adopted to create or maintain legitimacy with certain influential groups such as politicians and community groups (Crank, 2003). A critical aspect of institutional theory is isomorphism, which focuses more on why agencies change over time or become more alike. In 1983, DiMaggio and Powell introduced the idea that once organizations emerge as a field, they begin to become increasingly similar while 24 attempting to change. They also describe three isomorphic processes, coercive, mimetic, and normative. Coercive isomorphism is a result of powerful outside organizations, usually ones that the agency is dependent on for funds, pressure, or cultural expectations (DiMaggio & Powell, 1983). Mimetic isomorphism stems from organizational responses to uncertainty and normative isomorphism is associated with professionalization (DiMaggio & Powell, 1983). There is no indication of whether these categories are mutually exclusive or work together. While contingency theory and institutional theory are often described as opposing theories, some scholars suggest that they can be complementary, and that deeper understanding can be found when combining the factors. For example, Giblin (2006) finds that variables derived from both contingency theory and institutional theory were relevant and suggests the theories should be integrated. This dissertation does not seek to fully evaluate either line of theory but rather uses elements of each to exploratorily examine police use of social media, the modern role of the PIO and how organizational theory may inform understanding of this role. Police and media Police and criminal justice have been popular topics in multiple forms of media. From news media to popular television shows, the criminal justice system has been dramatized and popularized both negatively and positively. These representations are particularly important to understand as most individuals will not have personal contact with police, potentially leading to their perception of police and the criminal justice system to be based largely on these media representations (Surette, 1989). Diverse types of media may have varying importance and impact on public perceptions of police. For example, frequent viewers of police dramas have reported believing that the wealthy receive preferential treatment from police while frequent viewers of crime-solving shows do not share this belief (Dowler & Zawilski, 2007). While these effects may 25 seem of minimal concern, media perception can lead to profound consequences for police image. The impact of media on police image becomes even more pronounced when examining news media. Findings in the same study indicate that heavy consumers of network news are likely to believe that police misconduct is a frequent event (Dowler & Zawilski, 2007). Concerning traditional media, police sources may have little say in how this information is framed and distributed. As one of the most long-standing and important forms of media, news media in particular has attracted much attention from researchers attempting to understand media impact on perceptions of police and has the longest history. Within society, police must maintain their image and their legitimacy with the public. The importance of this relationship was highlighted even at the beginning of modern policing. Sir Robert Peel wrote in his “Principles of Law Enforcement” in 1829 that “the ability of the police to perform their duties is dependent upon public approval of police existence, actions, behavior and the ability of the police to secure and maintain public respect.” As policing has evolved, this principle remains relevant. In pursuit of this, police must be able to communicate with the public efficiently and effectively. News media provide police departments with access to a wide audience to distribute multiple types of information such as crime information or information about public safety (Chermak & Weiss, 2006; Ericson et al., 1989). Officers are expected to be morally good, effective crime fighters who work in the public’s interest as well as fulfill the department’s goals (Chermak & Weiss, 2005). At the same time, journalists and the press are working to provide the public with information, and if needed, hold government entities accountable for their actions. News media rely on police sources for accurate crime story information, while criminal justice organizations rely on news media to meet public demands for information (Chermak, 1994). 26 Through these relationships, criminal justice organizations control the amount of information they share and they are given a forum to respond to criticisms, generate support for programs, and further organizational objectives (Chermak, 1994). These relationships can become particularly important in times of emergency when police need to make mass notifications to a large audience in specific or broad areas. While these relationships are important to both police and media organizations, it is foreseeable then that the goals of each side may align at times and conflict at other times. Crime remains a popular news media topic, even experiencing a general increase in the percentage of news coverage since 2010 (Tyndall Report, 2019). As such, much research has examined how news media may impact perceptions of police over the past few decades. According to a meta-analysis focused on news media and perceptions of police conducted in 2018, the bulk of research surrounding news media and police focuses on high-profile incidents or negative news coverage (Graziano, 2018). Of these studies, the majority indicate some negative impact on perceptions of police from such coverage (Jefferis et al., 1997; Kaminski and Jefferis, 1998; Kochel, 2017; Lasley, 1994; Sigelman et al., 1997; Tuch and Weitzer, 1997; Weitzer, 2002). Consuming more news has also been observed to affect citizens attitudes about the guilt of officers involved in high-profile cases (Chermak et al., 2006). Beyond simply general effects, there has also been an exploration of predictors that are typically associated with attitudes towards police with the most consistently included predictors examined alongside news media were race/ethnicity, neighborhood context, and involuntary contact with police (Graziano, 2018). The coverage of high-profile incidents and negative police events seems to make up a substantial portion of media, but it is important to recognize that these may not represent the majority of in-person police encounters. 27 It comes as no surprise that a sizable portion of research examining police and news media investigates news coverage of high-profile or celebrated cases as this reflects what is being presented in the news itself. As was presented in Samuel Walker’s wedding cake model of criminal justice, the most media attention is placed on cases that are particularly gruesome or involve a famous person or persons (Walker, 2015). Celebrated cases make up the top tier of Walker’s model, representing the fact that these are the rarest cases, making up the smallest group/tier. These celebrated cases are unique in that they usually involve the full criminal process and receive an enormous amount of publicity, generally due to the nature of the crime (Walker, 2015). Finally, these small number of celebrated cases generally distort the public’s view of the criminal justice system as they represent a small number of the cases that make up the day-to-day operations of the system (Walker, 2015). The public may mistakenly make assumptions based on the media’s focus on these events, assumptions that are based on the few extreme cases rather than the most common. Further breaking down the research of celebrated cases, a substantial portion focuses on the use of excessive force against people of color (Graziano, 2018). These studies examine events such as a 1995 televised volatile arrest of a black teenage male in Cincinnati, the 1991 Rodney King incident, the killing of Malice Green in Detroit in 1992, and the 1996 beatings of two Hispanic males, among others (Jefferis et al., 1997; Lasley, 1994; Sigelman et al., 1997; Tuch & Weitzer, 1997). Some of these studies utilize well-established, regularly conducted surveys to capture attitudes before and after such events, while others that were underway when a high-profile incident occurred simply extended or manipulated their measures to capture effects (Graziano, 2018). Even across these varying events and types of research methods, results stay fairly consistent in finding negative impact across the majority of policing-related outcomes 28 (Graziano, 2018). These findings speak to the impact that single high-profile events may have on the police image. It is also worth noting that failure to find effects was generally in studies conducted outside of the US, which may be due to the generally high level of trust in police (Graziano, 2018; Kääriäinen et al., 2016). Beyond the examination of celebrated cases, there is another body of research that examines police use of media in a more routine manner or to promote certain programs or goals. These studies examine a different use of media as these stories are framing the police in a positive light rather than the negative coverage of police that celebrated cases typically center on. An example of this type of use may be the use of media to promote community policing efforts. The lack of research in this area may be due to the overall lack of representation of these efforts in news media. Typically, police department communication managers focus efforts, time, and resources responding to crime information incident requests (Chermak & Weiss, 2006). Crime incidents are generally considered high-priority by news organizations and community policing stories, when presented, are narrowly covered (Chermak & Weiss, 2006). The goals and history of community policing are not usually presented and contrary to the goals, community involvement is rarely encouraged in these news pieces (Chermak & Weiss, 2006). This example shows how the disconnect between the goals of the department and news media conflicts and may negate efforts. Police departments would benefit from diversifying their efforts and promoting such goals through alternative communication methods that do not rely as much on another organization or group’s goals. The internet, public service ads, target campaigns, community meetings, and social media may be different media where police departments may have more success in exposing community members to community policing efforts as well as any department initiatives that involve community participation (Chermak & Weiss, 2006). 29 While the body of police media relations research has spanned over the rise of the internet and social media, very few have considered the influence of news on these platforms. This gap in research is particularly interesting as in prior research, non-traditional media such as political talk shows have been shown to negatively impact perceptions of police (Moy et al., 1999). As the internet and social media would be considered non-traditional media, it is important to understand how findings may differ from those associated with mainstream media. Even so, mainstream media sources do exist on internet and social media so the untangling of the concepts may be complex. The few studies that do include internet and social media news sources differ in findings and measurement of internet use for news. One study measures internet news as the mean number of days an individual uses the internet to visit news sites and finds that this use was not consistently associated with views about crime or justice, yet there was a moderating effect of political ideology on this relationship observed (Roche et al., 2016). This study did not capture social media as a news source and only examined those who claimed to be using primary news sources online. Another study examined the self-reported most important news sources in a telephone survey and found that the internet was the most important news source for the greatest number of respondents (37 percent) and that those reporting local TV as their most important news source perceive police as more legitimate than those who report internet as their most important news source (Graziano & Gauthier, 2018). This measure did not contain any clear delineation of news from internet news site or social media but rather asked about the internet in general. The researchers also noted that coverage of policing on sites that allow for public responses (such as social media) could lead consumers to hold more negative views based on these comments, but this has yet to be explored (Graziano & Gauthier, 2018). 30 One more study examined attitudes toward police legitimacy and separated and compared reading news online and using social network sites. Findings indicated that only reading online news was significantly related to perceptions of police legitimacy and more time reading led to a less positive attitude about police legitimacy (Intravia et al., 2018). Regarding social media, male respondents who reported a greater use of social media perceived the police and being more legitimate (Intravia et al., 2018). While few studies specifically focus on social media, there is increasing support that it may have some connection to the perception of police. It should be noted that this sample was limited to a specific age group (mean = 19.74) and a specific group of individuals who were enrolled in criminology and criminal justice courses. The lack of research focusing on social media consumption seems like a major oversight as a 2018 survey found 34% of U.S. adults preferring to get their news online and a 2020 survey found 26% of U.S. adults claiming to get their news from the popular social media platform, YouTube (Geiger, 2019; Stocking et al., 2020). Internet news also has fewer filters than are applied to traditional sources such as TV or media as the internet is a host to a wide array of various and alternative views (Graziano & Gauthier, 2018). While this specific area of research is lacking, there is a fuller body of research that examines police use of social media at a more general level. Understanding the impact of social media and police department relationships with social media becomes even more important as local news is on the decline (Abernathy, 2018). As was previously discussed, police and news media relations may be complex but are important to both parties. These relationships may vary based on many factors including the level of the media. National news media have been noted by some police as being more aggressive and less trustworthy, while local media relations are typically stronger (Mawby, 2010; Surette & Richard, 1995). These relationships may be on the decline as local news experiences a fiscal crisis and 31 other platforms, particularly Facebook, rise as key infrastructural actors in the circulation of local information (Thorson et al.,2020). Public Information Officers Police communication has continued to widen in its professionalism and in its functions as department grow in use of various media types (Mawby, 2010). Prior to the 1980s, police departments were not likely to have formal media relations units or persons permanently assigned to such a position (Surette & Richard, 1995). To meet increasing demands for information and professionalization, many police departments began to employ Public Information Officers (PIO). The PIO is the gatekeeper of information from the department and is a specialist within the organization that is not only the public face of the department, but also dictates communication strategy (Chermak & Weiss, 2005; Skolnick et al., 1984; Surette & Richard, 1995). As PIOs rose in popularity and usage research has explored characteristics of PIOs and what the role of a PIO includes at different time points. One of the first studies to examine police PIOs sought to understand the demographics, job description, hiring process and attitudes of PIOs. This study was published in 1995 so it provides interesting insight into what the position was like at that time (Surette & Richard, 1995). It should be noted that this survey was only representative of Florida, but it is one of the first of its kind. The survey found that PIOs at the time were typically around 40 years old, three out of five were sworn officers and the majority had attended college (Surette & Richard, 1995). The respondents were mostly white males and they were found across agencies of various sizes (Surette & Richard, 1995). Interestingly, most of the PIOs indicated that they received no in- service media training until after they became PIOs and then they only received about 40 hours’ worth (Surette & Richard, 1995). Regarding their role, PIOs reported the most time-consuming 32 part of their job as fielding daily media inquiries, the second being news interviews and personal press conferences (Surette & Richard, 1995). Mostly, PIOs acted responsively and were not given time to put toward proactive publicity. Finally, PIOs were generally very satisfied with their jobs and reported positive relationships with local media (Surette & Richard, 1995). One noteworthy difference that became known in this study was the differences between sworn PIOs and civilian PIOs. While the groups did share attitudes and characteristics, they also differed in some ways that make them distinct. Sworn PIOs tended to be male, have a higher salary, and had more years of department employment while civilian PIOs tended to be females with lower salaries and were often hired specifically to be the PIO (Surette & Richard, 1995). While civilian PIOs were paid less, they appeared to work better with the media and held a higher level of job satisfaction (Surette & Richard, 1995). These findings were an important first glance into what department PIOs look like and what they do, for here research has expanded to include wider samples and observe differences in the role as new media arises and the job becomes more professionalized. Another survey of PIOs was conducted in 2000 and focused more on the public relations aspect of the role and organizational structure as it related to formalization of the PIO position (Motschall & Cao, 2007). Findings from this convenience sample survey indicated that the top three techniques used by PIOs were writing press releases, making formal and informal contact with media, and holding press conferences (Motschall & Cao, 2007). These findings further support that PIOs at the time were largely acting in a reactive nature by producing information and largely acting in a one-way flow of information out to news media. While this made up the majority of their work, the authors point out that PIOs are performing actions beyond this and 33 communicating with internal audiences as well as external audiences in other capacities (Motschall & Cao, 2007). One of the particularly interesting techniques that was reported by almost 75% of the sample was making informal contact with the public (Motschall & Cao, 2007). The authors suggest that this more recent diversity in techniques may be due to police departments moving towards a more open, service orientation. When the goals of PIOs were assessed, PIOs indicated that educating the public, giving the public facts so they will be supportive, and educating the public were important (Motschall & Cao, 2007). While the majority of the activities and the role of the PIO was still made up of reactive activities, there did seem to be indicators that PIOs were beginning to use their position more proactively. They valued a more comprehensive approach to the position and saw themselves as an important liaison between the department and the public. This study also explored organizational aspects that may impact the PIO position. While the findings indicated more formalization of the role, staff sizes, and budgets did not seem to be increasing (Motschall & Cao, 2007). While police departments and leaders saw the importance of the PIO position, resources were lacking for the position. Budgets for PIOs were severely limited with most reporting no budget allocations beyond equipment and training (Motschall & Cao, 2007). It seems that even as this position expands, departments are limited in their resources (time, money, training, etc.) and not likely to allocate them to PIOs. The reasons why this is occurring warrant further empirical investigation. To expand on the baseline information known about PIOs, a national survey examined PIO activities and their proactive publicity efforts was conducted by Chermak and Weiss (2006). This survey combined a survey of police PIOs with surveys of media organizations in order to better understand the relationships between the two. The findings of this research supported prior 34 research in that the police-media relationship was symbiotic, with nearly 90% of the PIO agreeing or strongly agreeing that the current status of their relationship with most news organizations was good (Chermak & Weiss, 2006). Even so, it would be incorrect to assume that this relationship is constantly at this state. There are surely times when the organizational goals conflict and the relationships weaken, particularly when police are faced with high profile cases or corruption (Chermak & Weiss, 2005). PIOs are expected to maximize the positive images of the department and minimize the negative images that are presented in the news (Chermak & Weiss, 2005). It seems that the benefits the police organizations and news media gain from one another seem to keep each organization motivated to maintain these relationships for the most part but at times these relationships are strained and opposing. Both sides recognized that while they considered their relationship with the other positive, there was possibility for these relationships to flip quickly (Chermak & Weiss, 2005). Similar to prior studies, the majority of PIO activity was spent responding to media requests for information and only about 15% of the PIOs’ time was spent on requests about police initiatives such as community policing programs (Chermak & Weiss, 2006). While news personnel did believe that the public was interested in community policing stories, relatively little effort was put towards these stories (Chermak & Weiss, 2006). The media provide an opportunity for the police to shape the views of the public and promote programs and policies, yet this use was rarely capitalized upon by the departments in this study. While these findings were not surprising, they did help to confirm that the results of other more limited studies were observable at a national level. Other findings from this study were centered around publicizing community policing. The majority (96%) of the agencies that were in the sample had some type of community 35 policing program and about 80% of PIOs thought that their department did a good job publicizing positive aspects of the agency (Chermak & Weiss, 2006). On the other hand, only about 40% of the media sample agreed that the police did a good job of keeping the media informed about community policing (Chermak & Weiss, 2006). Overall, it was concluded that law enforcement agencies should consider attempting larger marketing strategies to promote community policing, PIOs should play an important part in this development, and that police may need to consider platforms outside of news media for these reasons (Chermak & Weiss, 2006). Updating understanding of PIOs Much has changed in policing and media since the majority of the research surrounding PIOs was conducted. PIOs have become even further professionalized and professionalized in the past few decades. The National Information Officers Association (NIOA) was created in 1989 and held their first conference in 1991 with forty members. Since then, the NIOA has continued to flourish to be made up of hundreds of members, include ten different regions throughout the US and Canada, and publishes a bimonthly newsletter (NIOA, 2020). The continued growth and success of this organization indicates the continued growth and professionalism of the position. While this position has continued to grow, little research has reevaluated PIOs and the role they play. One study used more recent data to update the understanding of “police spokesperson” demographics in 2019. This study created a database of police spokesperson rank, status, gender, race, and age based on publicly available data and subjective characteristics coding (Simpson & Wetherell, 2020). Similar to the demographics recorded by Surette and Richard more than two decades prior, the majority of the spokespersons were white, sworn, males and were around 40 36 years of age (Simpson & Wetherell, 2020). The males spokespersons were also more likely to be sworn than female spokespersons (Simpson & Wetherell, 2020). These findings are intriguing as it was expected that departments would be working to diversify their representation. This study also brought to light the informal finding that most agencies explicitly charge their PIO with managing the department's social media as this is where most of the information about the spokespersons was found (Simpson & Wetherell, 2020). While this study is interesting, it is limited in its scope and uses subjective coding of demographics rather than officially reported data. Also, while updating demographic information about PIOs in recent years may be interesting and important for representing department diversity, there is much more regarding police communication strategies that warrants further investigation. It is clear that the majority of PIOs time and efforts in prior decades were spent responding to news media requests for information (Chermak & Weiss, 2006; Motschall & Cao, 2007; Surette & Richard, 1995). These roles may have changed as police community relations change as well as new media is introduced. While the internet and social media were present when the more recent studies were conducted, there has been little acknowledgment or examination of how these media have altered the role. Particularly in present times, as local media is in a decline and social media platforms are replacing these media, it is imperative that further research examines the impacts of this shift. The roles and responsibilities of a PIO in a police organization have surely changed with the times. Police and social media While media, in general, are well known to have an impact on perceptions of police (Chermak et al., 2006; Dowler & Zawilski, 2007), little research has focused on social media specifically. While police agencies have reported using social media prior to 2006, one 37 nationally representative sample of police agencies found that the majority (about one in five agencies) began using social media as part of their official operation in 2012 (Kim et al., 2017). There is potential that the use of social media by police has eliminated the information intermediary, allowing police departments to communicate to their audience quickly, in a more interactive way, and outside the constraints of news media formats. Social media as a form of communication may also come at a lower cost for the department and provide benefits that are not typically available with traditional methods of police-community communication. The differences in police information dissemination via social media compared to traditional media warrant more systematic research. Social media is clearly used by police departments as the IACP annual survey on social media found 96.4% of responding departments used social media in 2015 (IACP, 2015). Even so, little research exists focusing on questions such as why, when, and how the police utilize this new communication form, and what variables explain differences in activity and engagement. The question of how police use social media from an outside perspective has been explored through multiple content analyses of posts. These studies are limited to data sets gathered from social media platforms, mainly Facebook and Twitter. In examining content, studies have found that police departments use social media to push out information to the public they serve, rather than “pull in” information (Lieberman et al., 2013; Meijer & Thaens, 2013; Dai et al., 2017). Pushing out information refers to the department simply posting something informative with no expectation for interaction or reaction from the audience. Pulling in information would include posts asking for some response such as asking for information from the public about an event or asking community members to call in if they have any tips. Even so, there is some evidence that this is changing or not the case in all departments. More recent 38 research has found that some departments use a pull strategy to solicit information from the public even more often than they push out information (Hu et al., 2018). The differences in strategies warrant more examination as departments may be growing in their understanding of social media, or they may be changing their strategies in response to some other factors. While there have been multiple analyses of content created by police and posted on social media (Hu et al., 2018; Lieberman et al., 2013; O’Conner, 2017), little research has examined the strategies used by police on these platforms. The few studies that do examine social media and website strategies utilized by police departments generally gather data from the platforms rather than the producer of the content (Dai et al., 2017; Rosenbaum et al., 2011). The motivation behind the posts and content created has not been explored. This limits the findings as far as a deeper understanding of the intent and creation of the messages by police. Large gaps remain in this literature including an understanding of the role of the PIO in department use of social media and what predicts their outputs in this space. There is evidence that police chiefs in the U.S. believe they are using social media sites in ways that are traditionally linked to the PIO position. In the 2016 IACP survey the top reasons for using social media included the following: 91% use it to notify public of safety concerns, 89% use it for community outreach and engagement, 86% use it for public relations and 72% use it to monitor public sentiment (Kim et al., 2017). Further, the top two most valuable training respondents wished for were “strategies for engaging the community” (25%) and “guides on improving agency use of various types of social media” (23%) (Kim et al., 2017). Due to the impact that communications sent out on social media may have, it comes as no surprise that the most common manager of social media accounts in police departments is the public information officer (PIO) (Kim et al., 2017). This IACP survey is one of the few large-scale studies that has 39 been conducted which examined police and social media at a national level and from the perspective of the department themselves. Little research has moved beyond the content of social media to further understand the purpose and potential of the media for PIOs, as well as what explains variations in social media activity and engagement. Questions remain regarding goals of department social media use aligning with the actual use of the social media pages. The perspective of police PIOs has not been captured up to this point and may provide valuable insight on the actual use and understanding of social media. Organization characteristics and social media Organizational characteristics and use of social media have been explored through examination of different government organizations’ use of social media. These studies attempt to focus on determining factors of strategies and policies that these organizations utilize. Strategies appear to be motivated by both internal and external factors (Medaglia & Zheng, 2017). Local governments who have used social media longer tend to utilize more advanced features and municipalities with more higher-educated people tend to use social media more interactively (Faber et al., 2020). In examining police agencies specifically, internal factors that determine government social media strategy include the structure of the organization, with formalization, hierarchy, and centralization of the organization being reflected in the social media strategy (Meijer & Thaens, 2013). Structural government characteristics have also been found to hinder police use of social media. Barriers are created when resources are limited to devote to implementation and maintenance of a professional web-presence (Dekker et al., 2020). Structural government barriers also created difficulty in embedding social media into existing organizational processes as well as the management of social media (Dekker et al., 2020). Dekker et al. (2020) found 40 barriers to include a lack of clear protocols, specifically, a need for guidance in how often to post, what platforms to use, and who should be responded to online. There are also cultural barriers identified by research that impact police use of social media. It is generally difficult for police organizations to maintain effective communication due to the constantly changing social media landscape and lack of adaptability to fitting communication to each platforms affordances (Procter et al., 2013; Dai et al., 2017). Police agencies also tend to be concerned about risk of the heightened visibility when utilizing social media (Dekker et al., 2020). Agency size has been connected to some aspects of social media management, with smaller agencies managing their social media presence with a wide range of staff and larger agencies (those employing 100 or more full-time sworn personnel) found to be more likely to delegate social media management to PIOs (Kim et al., 2017). Interestingly, Larger agencies are also more likely to use informal tones with 33% of large agencies reporting always or almost always using informal tones (Kim et al., 2017). In contrast, how long the agency has used social media was not found to be a significant factor in determining how they use the technology (Kim et al., 2017). PIO management of social media As was previously discussed, social media provides police with a new method of communication with the public, one that allows for even more control of the message being created. PIOs no longer have to rely on news media to reach a large audience and they are able to push any messages they wish without the editing and curation of news media. PIOs are also able to potentially use social media to gain instant feedback about the stories they choose to share, as well as respond to this feedback quickly. Content analyses of police use of social media indicate that while these potential new uses exist, it was common for police departments to utilize social 41 media as they did traditional media by simply pushing out information (Dai et al., 2017; Lieberman et al., 2013; Meijer & Thaens, 2013). Yet it seems that as understanding and use of the platforms grow, there may be more diversity in this use including more soliciting of information and interaction (X. Hu et al., 2018). Still, there is little research that actually focuses on the content creators of social media, likely the PIOs. The few studies that do explore PIOs relationship with social media are limited to extremely small sample sizes of 25 and 13 (Graham, 2014; Hughes & Palen, 2012). Both of these studies involved in-depth interviews of PIOs in limited areas and from local government agencies, not specifically police. These analyses are limited but may be useful in providing a base of information on the topic. In discussing social media, PIOs brought up important issues such as barriers, dialogue promotion, engagement, and control over message (Graham, 2014). Barriers to social media uses were discussed in findings from both studies. PIOs discussed difficulties with getting approval creating slow turn around, lack of money, time, and staff (Graham, 2014; Hughes & Palen, 2012). Budget concerns were specifically highlighted when PIOs discussed getting involved in more demanding platforms such as YouTube (Graham, 2014). Another barrier brought up was the fact that the technology was still fairly new to the organizations and some leaders remained skeptical of the benefits and potential of social media (Graham, 2014). Beyond the barriers mentioned, many PIOs in these studies also discussed the potential benefits and uses of social media for their organization. PIOs reported using social media such as Facebook to send incident updates, thus leading to a reduction in media inquiries (Hughes & Palen, 2012). In fact, some PIOs brought up the notion that the public in some cases are the first “reporters” to the scene and are able to post video and images before news media arrives 42 (Hughes & Palen, 2012). This could potentially be an issue if these citizen reporters are pushing out information that is untrue. PIOs also saw social media as a more direct link to the public and as a tool to help promote two-way dialogue (Graham, 2014; Hughes & Palen, 2012). Yet, PIOs also felt as if they now became reporters, a role that they were perhaps not ready to take on (Hughes & Palen, 2012). Even so, PIOs generally recognized the freedom that social media brought to their job, allowing them to push out exactly the information they want at exactly the time they want (Graham, 2014). Further, the authors point out that this freedom also creates the need for PIOs to better understand and connect with their audience in a way that they were not required to before social media (Hughes & Palen, 2012). This need increases the need for appropriate education on the use of social media and what tools work best for what goals. In regard to education, departments seem to be most interested in using social media to engage the community and improving their social media use over all as these were the top two most requested trainings in the IACP survey, making up 54% of the respondents (Kim et al., 2017). A PIOs use of social media may depend on not only their education of the platforms, but also on other personal characteristics of the individual. While the PIO is acting as the face of the department, there is room for their own individual strengths and weaknesses to affect how they use social media. Especially as social media policies were limited when some PIOs first began using the platforms. In 2010 an IACP survey found that only 35.4% of surveyed agencies had a written social media policy, in 2016 that portion increased to 80%, yet there is evidence of varying degree in the quality of police social media policies (Kim et al., 2017; Rodgers, 2016). Without rigid policy dictating the use of social media, there is much room for PIOs to interpret and decipher how social media should be used in the name of the department. There is little research that addresses this information directly as there is little research that focuses on the 43 producers of police social media content in general. Research regarding the demographic makeup of police spokespersons however indicates that the spokespersons are overwhelmingly white, middle-aged, sworn males, which may lead to little variation in use in relation to demographics as there may be little variation in demographics (Simpson & Wetherell, 2020). Due to lack of variance and little empirical evidence, the most important personal level factors may be the PIOs understanding of social media itself, training with the platform, and motivation of use. Relations between the news media, or the press, and police have been examined since the introduction of social media in the Australian and British context more recently (Colbran, 2020; Lee & McGovern, 2014). These contexts are quite different from the United States but findings from these studies suggest that police “press officers” are now able to potentially bypass traditional media through social media (Lee and McGovern, 2014). In Australia, it was argued that the police maintain access to critical data which they may be less likely to release to journalists for fear of being portrayed in a poor light when they may release the information themselves through new media as they wish and when they wish (Ellis and McGovern, 2016). In Britain, police similarly mentioned the idea of bypassing media to promote more positive stories that traditional media was less likely to publicize (Colbran, 2020). While these findings are in different contexts that each have their own complexities to consider, it is expected that this study may find similar sentiments. As police PIOs in the US see the same potential to use social media as a way to bypass their relationships with news media, relationships may become more strained. While studies exploring the intersection of PIOs and social media are limited, it is clear that PIO work has evolved and changed with the introduction of social media. The role of the PIO may include new tasks, new responsibilities, and even a new language. Social media has 44 introduced fresh complexities along with many potential benefits and barriers to the job. While change is clear, there is an extreme lack of nationally representative research surrounding the topic. Knowledge of the role of the PIO is generally restricted to their relationship with traditional or news media. With social media continuing to rise in popularity and even stepping in and taking the role of local media in some areas this knowledge is vital (Thorson et al., 2020). Understanding of the PIO and social media can no longer be delayed. Little is known about the strategies PIOs use on social media, the department resources (training, time, and money) allocated to social media, if these differ from what is used in traditional media and the individual and organizational level variables that may influence these. These issues are even better highlighted in an examination of two high-profile events of 2020, COVID-19, and the death of George Floyd. PIOs and high-profile cases As was discussed previously, celebrated, or high-profile cases can have detrimental effects on police image and image of the criminal justice system as a whole. These highly publicized events do not represent the whole but rather a part of the day-to-day actions and events in the criminal justice system. Police reactions to noteworthy events are extremely important in the public eye and can shape the image of a department. There have been multiple events, even in the past decade, which have been so high- profile that they attracted attention from media, the public, and researchers. These events have transcended the specific geographic spaces they take place in and represent a general level of discontent with policing in the US (Nix & Wolfe, 2017). A few of the most impactful events took place in the summer of 2014. Shortly after Eric Garner was killed by police officers in New York City, Michael Brown was fatally shot by police in Ferguson, Missouri. These cases both 45 involved a white officer and a Black, unarmed citizen. Media surrounding these cases spread far and fast as people across the nation reacted to and began to question police use of force. In fact, the news of Michael Brown’s death emerged on Twitter before reaching mainstream media and generated more than 10.6 million tweets in the 8 days that followed the shooting (Hitlin & Vogt, 2014). As new surrounding this case continued to spread on social media and new media platforms, the discourse expanded to include police brutality against people of color and systematic racism in America (Cobbina, 2019). Social media played a large part in mobilizing the protests that followed as well. In interviews of protesters in Ferguson, 39% discussed the role of social media played in their efforts (Cobbina, 2019). Social media was an important tool for those on the ground to share information with those outside of the movement and as a means to quickly exchange information (Cobbina, 2019). Social media was an important form of media for many people involved on multiple levels. Information about this celebrated case particularly was spread through many different people and social media platforms, emphasizing the pervasiveness of this media type in modern times. Since Ferguson, researchers have become interested in what has become known as the Ferguson Effect. The Ferguson effect refers to officers becoming less proactive in their actions as a response to increased scrutiny of law enforcement, which was theorized to cause an increase in violent crime rates (Nix & Wolfe, 2018). There is some support for parts of this “effect” as in the years following Ferguson, officers did feel that there was a “war on cops” (Nix et al., 2018). These high-profile events and the media coverage surrounding them, particularly in Ferguson, have been found to impact not only the public perception of police, but also police self- legitimacy. Negative publicity surrounding the police in general has been found to have an impact on individual officers’ sense of self-legitimacy (Nix & Wolfe, 2017). Specifically, 46 officers who were less motivated to do their job as a result of negative publicity were significantly less likely to believe they were legitimate authority figures, even when controlling for other known predictors of self-legitimacy (Nix & Wolfe, 2017). Low self-legitimacy in officers can cause de-identification with the agency and it’s goals and may cause officers to pull away from interactions with community members (Nix & Wolfe, 2017). This finding is important because it underlines the impact of negative publicity and high- profile events that reach far beyond the jurisdiction in which they take place. Another important take away from this study is that perceptions of organizational justice remained the strongest predictor of self-legitimacy, suggesting that agencies may still cultivate confidence in their officers through internal efforts (Nix & Wolfe, 2017). These findings underscore the importance of police organization response to high-profile events. Police organizations must find effective ways to respond to high-profile, negative events as these events have the potential to impact not only the public, but also may negatively impact officers. How police organizations respond to “legitimacy deficits” is critical in bridging the gap between what the organization believes about their legitimacy, and what the audience perceives (Bottoms & Tankebe, 2013). PIOs are an important instrument that departments may use as they are intended to bridge the knowledge gap between the public and the organization. Line level officers are not the only ones to feel the influence of celebrated cases. In further investigation of the management level officers, those who felt their communities perceived police as legitimate were less sensitive to recent high-profile police shootings (Nix & Wolfe, 2018). This is particularly important as police legitimacy relies on not only considering what makes the public view the police as legitimate, but also what the police believe about how the public views them (Nix & Wolfe, 2018). The finding that the Ferguson effect is real among 47 management level officers is troublesome as they are in the position to impact organizational justice, which, as was discussed, is important to officer self-legitimacy. The authors call for police managers to create strategies to intervene in this process and prevent the decay of community trust (Nix & Wolfe, 2018). Such strategies would fall on PIOs who are generally responsible for releasing information from the department and acting as the face of the department. While PIOs are constantly responding to smaller events, their response to major, high- profile events are particularly important for all parties involved, the police and the public. Effective communication between the police and the public is essential to the police function, and this is particularly emphasized during and after celebrated cases. Research has highlighted the complexities of how policing is affected by events, even when they do not occur in the same geographical area. The year 2020 brought two high-profile events of vastly different natures that surely impacted police and demanded some action on the part of department PIOs, COVID-19 and the killing of George Floyd. These events brought their own complexities and nuances but the research on prior celebrated cases informs this study and underlines the importance of considering these events and how they shaped policing and the job of the PIO this year. For review, each of these events will be briefly summarized. COVID-19 SARS-CoV-2 (COVID-19) virus was officially declared a pandemic on March 11th, 2020, by the World Health Organization. A pandemic of this proportion is an unprecedented modern health crisis which has presented new challenges for individuals and organizations. Since this time, the COVID-19 outbreak has continued to spread and continues to take up much of the news media’s attention. The pandemic served as a reminder of the importance of local 48 news as residents in cities and rural communities searched for reliable, comprehensive, and accurate information on what is occurring in their local space (Abernathy, 2020). All levels and types of news media have covered the topic, almost on a daily basis, tasked with keeping the public up to date on an event that is continually changing. In addition, as the world adapts to the pandemic, police departments have been impacted on many levels. Police organizations have been particularly affected as they have been asked to enforce revised local or regional health ordinances such as travel bans, mandatory quarantines, public gathering limits etc. (IACP, 2020). Special considerations have been brought to light as departments decide how to protect themselves while continuing to carry out their duties, which are being altered to fit the pandemic. Beyond that, the operational landscape of policing has also been impacted as police take their role as first responders to medical emergencies and are potentially exposed to COVID-19 themselves (Lum, et al. 2020a). In an effort to understand how the pandemic continues to influence policing in the U.S. and Canada, the IACP and the Center for Evidence-Based Crime Police (CEBCP) partnered on a multi-wave panel survey of police agencies. Findings from the first wave (March 23rd, 2020) indicated some major shifts in policing including the following: 91% of agencies had written criteria or guidance on changes to officer responses to calls for service due to COVID-19, 43% of responding agencies had stopped or meaningfully changed their response to at least twenty percent or more calls for service, 76% of respondents provided their officers with formal guidance to reduce physical arrests for minor offenses, 73% had adopted policies to reduce or limit community oriented policing services and 39% had increases community presence for certain locations such as grocery stores and hospitals (Lum, et al. 2020a). Agencies were also asked to share ideas their agencies used to successfully mitigate challenges due to COVID-19. 49 Agencies stated that one strategy was to increase communication with police department employees and community members daily (Lum et al., 2020a). These agencies claimed that using various forms of communication seemed to improve morale and reduce false information (Lum, et al. 2020a). The second wave of the survey, which was completed May 2020, included all previously asked questions as well as additional questions aimed at better understanding key issues which arose from wave one. One key issue that was further clarified in the second wave of questions focused on what categories of calls police were seeing differences in, rather than grouping them all together. The calls for service were lower in April 2020 as compared to April 2019, but increases were seen in calls for domestic violence (42% reporting an increase) and calls related to people in mental distress (47% reporting an increase) (Lum, et al., 2020b). COVID-19 modified agency operations with 95% of responding agencies reporting providing patrol officers with formal written criteria specific to COVID-19 and how to respond to calls for service and 85% had provided formal guidance to detectives on how to modify their investigative activities as well (Lum, et al., 2020b). Substantial reduction in use of arrest and jail intake was reported and agencies continued to indicate reduction in proactive and community engagement activities (Lum, et al., 2020b). Paying particular attention to communication during COVID-19, 27% of respondents reported an increase in their use of social media because of the pandemic and 79% rated themselves as “Excellent” or “Good” in their ability to answer the community’s questions and concerns about COVID-19 with confidence (Lum, et al., 2020b). These findings suggest that policing has been impacted in a myriad of ways by the global pandemic. It is likely that these numbers will continue to fluctuate as police departments continue to adapt to the pandemic. 50 George Floyd On May 25, 2020, a 46-year-old Black man by the name of George Floyd died in police custody after being arrested by a white Minneapolis police officer, Derek Chauvin. The arrest footage caught on a cell phone camera as well as body warn cameras of officers at the scene shows the police officer kneeling on Floyd’s neck for nearly 8 minutes (State of Minnesota v. Chauvin, Compl. ¶ 8-11, 2020). The cell phone video went viral after being posted to Facebook and the four officers involved in the incident were fired the next day on May 26th. Even after the officers’ termination, protests began to rise in the streets of Minneapolis and extended to cities across the country. Chauvin was charged with murder and manslaughter on May 29th, yet protests continued and even sparked to an international level. Related protests and anti-racism demonstrations took place in Australia, France, Germany, Spain, and the UK in the weeks following (George Floyd Death: US Protests Timeline - BBC News, 2020). As these events unfolded, PIOs were tasked with responding to community members demanding change. PIOs had to step up and act as the face of not only their organization, but policing in the U.S. The Minneapolis Police Department’s own PIO was criticized for not being clearer about what happened during Floyd’s arrest and the city council voted to move the PIO position out of the police department and into city hall communications (Minneapolis Police’s Public Information Officer To Be Under City’s Supervision – WCCO | CBS Minnesota, 2020). This move caused concern among journalists, police, and PIOs both in and outside of Minneapolis as it may set a precedent that could lead to less transparency, lack of timely release of information, and increase political influence in messaging (NIOA, 2020). The society of professional journalism also noted that an effective PIO must have trust of both the police and 51 journalists, which takes time that they did not believe a PIO based in city hall could dedicate to the job (WCCO, 2020). Beyond departmental concerns, PIOs are also individuals who had to absorb the impact of these events along with the continuing to respond to the ongoing pandemic. In a public letter addressed to PIOs and posted to Twitter and the NOIA website on June 3,2020, the president of the NIOA, Ashley McDonald, stated the following: “On a personal level, we watch our family members, colleagues, friends, and even strangers suffer by either words or actions. Our law enforcement friends. Our friends of color. Our law enforcement friends of color. Words hurt. Actions hurt. Lack of action hurts. We cannot hide from it. We’re embedded. We’re engaged in social media, and those social media posts can be ruthless.” There is good evidence to indicate that George Floyd’s death made a difference in the work of the PIO, which warrants further investigation. Hypotheses While there are exploratory questions that could be asked, such as what contingencies influence PIO use of social media or what is the motivation of PIOs to use social media; there are also hypotheses that may be proposed based on prior research that may be able to identify specific influences. These hypotheses connect individual, organizational, and combined factors that have potential to impact police organization use of social media, particularly department activity (number of posts) and citizen engagement. Department level hypotheses are based on organizational theories including contingency and institutional from prior literature. 52 Individual level hypotheses: H1: The amount of media training will be positively related to both department activity and citizen engagement. H2: The length of time an individual has used Facebook will be positively related to both department activity and citizen engagement. H3: PIOs who feel competent in their use of social media for the department will be positively related to both department activity and citizen engagement on Facebook. H4: PIOs who feel Facebook is important for their department will be positively related to both department activity and citizen engagement on Facebook. H5: PIOs who enjoy using Facebook for their department will be positively related to both department activity and citizen engagement on Facebook. Organizational Level Hypotheses, Contingency: H6: Heterogeneity of the community served will be positively related to the department’s use of social media. H7: Size of the agency will be positively related to both department activity and citizen engagement on Facebook. H8: Command Staff support will be positively related to both department activity and citizen engagement on Facebook. H9: Amount of department resources (budget, time spent on training, technological support) devoted to social media will be positively related to department activity on Facebook. 53 Institutional: Isomorphism, Coercive: H10: Agencies who have received external funding for departmental social media activities will be more likely to use social media more frequently. Normative: H11: Accredited agencies will use social media more frequently. H12: PIOs with more formal training will use social media more frequently and have greater citizen engagement. H13: PIOs who belong to professional organizations will use social media more frequently and have greater citizen engagement. H14: PIOs who seek outside sources of information on how to utilize social media will use social media more frequently and have greater citizen engagement. Mimetic: H15: PIOs who regularly monitor and/or model their departmental social media use after other law enforcement agencies will use social media more and will receive greater engagement. Conclusion While there is research on police PIOs, there are major gaps to fill, due to the introduction of recent technology and lack of recent inquiry. Most of the research that composes our knowledge of the role of PIOs is now dated and does not reflect or investigate PIO use of social media. The role of the PIO for the past decade has remained largely unexamined and with the introduction of technology and new media, not only has communication changed but relationships with traditional media that used to be important may also differ. A better understanding of the current practices, the role of social media, and the impact of current events 54 is required. This study represents a solid step in filling these gaps by combining multiple data sources, a national level survey, social media data, and department information. Due to the timing of this study, there was also the opportunity to capture response to recent, significant events. 55 CHAPTER 3: METHODOLOGY This chapter will begin by discussing the survey data collection methods used, including the benefits and difficulties. Then, the social media data collection will be presented including the use of the CrowdTangle for data collection. Finally, this chapter will discuss the two high profile cases that are of interest to the study and the analytical plan. Data Collection The study consisted of the following two data gathering techniques: an online, self- administered survey of randomly selected police department PIOs and a collection of public social media data. The survey measured the role of the PIO with items adapted from Chermak and Weiss (2006), along with newer items that were developed specifically to capture a) methods used to market police departments on social and traditional media; b) the personnel responsible for this work; c) perceived benefits and drawbacks of using social media; and d) training and knowledge of various social media platforms (see Appendix A for survey items). These data were then attached to a collection of agencies’ publicly available Facebook data and department characteristic data. The PIO survey was administered in three waves to a random sample of 582 police PIOs serving populations over 50,000 or employing at least one hundred sworn officers in the U.S. The sampling frame was created from the 2016 Law Enforcement Agency Roster (LEAR) produced by the Bureau of Justice Statistics (BJS). This sample was chosen in order to include police departments most likely to have a PIO who also manages social media, yet also allow for a range of department sizes. The 2016 IACP survey found that PIOs were the most reported individual responsible for social media in agencies serving populations 50,000 and above (Kim et al., 2017). The resulting random sample was made up of 4.2% state police, 59.8% local police 56 and 36% sheriff’s departments. Unfortunately, the organizations surveyed are not representative of police departments serving smaller populations in the U.S. but do represent agencies that are most likely to have a PIO who has experience using both social and traditional media. The police PIO questionnaire and consent form were sent electronically through Michigan State University’s Qualtrics system to each PIO. To determine the correct and most appropriate email address, I identified the PIO through department websites, department social media pages, or news releases put out by the department. Assuring the correct email address is on file is important for the initial survey and also allowed for more direct follow-up if needed. If I was unable to identify the PIOs direct email, an email was sent to the department or chief asking for the information to be forwarded on to the PIO or person responsible for department media. Some agencies did not have public emails at all, but instead had a message system built into their website. If this was the case, I sent the message through the department’s site system, asking it to be forwarded to the department PIO or person responsible for department media. The topics of improving use of social media and engaging the community have been reported by police departments as the two most valuable training topics, as seen in the 2016 IACP survey (Kim et al., 2017). Based on this finding, it was expected that this topic would be of high interest to PIOs currently. This notion was confirmed during the pilot and after the launch of the survey with several individual PIOs reaching out to thank me for covering this topic. Even with this expected interest and PIO survey studies receiving high response rates in the past (Chermak & Weiss, 2006), it was critical to take steps to improve the legitimacy of this survey. Several methods were utilized to encourage responses. I administered the survey at Michigan State University (MSU) and through the MSU Qualtrics system, and from an MSU email. By identifying the PIO through online sources, surveys were able to be addressed to specific 57 addresses, rather than guessing or addressing it to the department in general. The questionnaire was also kept as brief as possible yet was still expected to take about thirty minutes to complete. PIOs were told the expected time up front so that they could make the decision if and when they would take the survey, participants were also allowed to return to complete the survey if they could not in one. The majority of the questions were close-ended and provided in a convenient online format. Finally, interest in further understanding of social media and police community relations is important to PIOs. The survey was designed and administered in three phases. Since the survey included added items that had not previously been tested, the survey was pretested with three current, former, or recently retired PIOs. These individuals were asked for their feedback to identify problematic questions and for insight into any other improvements. After the pretest phase, the survey was sent out to the study sample in a few phases. Two weeks after the initial email, a reminder email was sent to those who have yet to respond, reminder emails were sent about every two weeks with a total of three reminders. While ideal timing for web surveys varies depending on the population, adequate time must be given for response and time should be limited enough for the initial contact to be remembered (Dillman et al., 2014). Using a different mode for follow-up contact would have been ideal as it eliminates some of the issues with email contact such as the survey invitation being sent to spam (Dillman et al., 2014), but it was extremely difficult to reach this population by phone. Many phone numbers were limited to media only or required messages to be left which did not prove fruitful for follow-up. Several departments also no longer provided direct numbers to individuals such as the PIO as they requested to be reached by email instead. As such, emails served as the primary form of contact for this study. Qualtrics has built in features that were able to catch bounce back emails and I 58 was able to update these failed emails as best as possible. Some of the emails could not be redirected to a successful address as the contact information may have been out of date. The following sections were included in the survey: Introduction, PIO role, social media, Current Events, and Demographics. Appendix A includes the full survey instrument. The PIO role section of the survey included many items adapted from Chermak and Weiss’ (2005) original PIO survey. These items were crafted to capture information about what current communication methods are being used by PIOs, what strategies are being used by PIOs, and what the relationships between PIOs and news media are now like. Items were altered to include and capture new media (social media) responses as well. The social media portion of the survey included items that were developed based on some items from Chermak and Weiss’ (2005) prior survey, prior research, and my own experience in prior qualitative interviews with police who create their department’s social media content. These items attempted to capture the extent of social media use by the department, how this differs from traditional media, what department resources (money, training, and technology)are being put towards use of social media, how PIOs work with other units to share information via social media, and what social media experience and training the PIO has through both the department and their own personal use of social media. The final portion of the survey focused on the current events of COVID-19 and the death of George Floyd. Items in the concluding section were created specifically for this survey to capture how each of these events have impacted the PIO’s strategy overall and their use of social media specifically. The questions utilized are general to allow for interpretation and help build a solid base understanding of how these celebrated cases impacted PIOs. There were questions that 59 asked how each event impacted relationships with media as well as use of social media in their positions. Social Media Data Collection While the PIO questionnaire provided valuable insight into social media strategies, the actual use of social media was captured by social media posts. An agency’s public Facebook data was captured using a CrowdTangle Dashboard. Facebook was chosen as the platform of focus due to its popularity among police departments of multiple sizes and its longevity. Data from the 2015 IACP survey indicates that the platforms most used by police departments were Facebook (94.2%), increasing the likelihood that the sample will use this platform. Facebook data also allows for more rich investigation as data can capture multiple types of media and reactions. Facebook allows for multimedia posting with customizations such as videos, pictures, music and now even provides several types of reactions from audience members (likes, loves, laugh, mad etc.). Few studies have examined interactivity of users and police (or any government organization) social media or the motivation behind posting to department social media (Medaglia & Zheng, 2017). This study partially fills this gap through collection of social media data metrics and post content. Social media account specific data were collected, such as the post content, number of posts in a specified time range, and reactions. This study used social media metrics proposed by Bonsón et al. (2017) and Haro-de-Rosario (2018) to measure department activity and dimensions of citizen engagement. Department activity is defined as the average number of posts by department per working day (Bonsón et al., 2017). Citizen engagement is measured through an index made up of the following three metrics: popularity, virality, and commitment (Bonsón et al., 2017; Haro-de-Rosario et al., 2018). Popularity (P) is measured as 60 the average number of likes per individual post. Virality (V) is measured as the average number of shares per post. Commitment (C) is measured as the average number of comments per post. To compute the final engagement index (or interactions per post), popularity, virality, and commitment were summed (Bonsón et al., 2017; Haro-de-Rosario et al., 2018). The equation for the index is as follows (𝑃 + 𝑉 + 𝐶) = 𝐸𝑛𝑔𝑎𝑔𝑒𝑚𝑒𝑛𝑡. The measures of activity and interactions represent the outputs of the PIO on each platform and are linked to the inputs from the survey data. Interaction per follower at the time of post was considered but ultimately could not be used for this sample as there was a large portion of data (followers at time of post) missing. Due to this being a slight change from the original authors work, who did include followers at the time of post, the engagement index will be referred to as interactions per post for this dissertation and analysis. While these are not perfect measures of reach, as many more people may see a post and not be a follower or react to the post, these measures do give a sense of the audience that has potential to see posts regularly and those who choose to actively engage with the department. Linking social media and survey data allows for comparison of department social media outputs and PIO reported inputs such as time, training, and budget. Examination of police social media data has, up to this point, largely focused on content analyses of posts. The method of linking survey responses to publicly available digital data has been previously used in research of public actors such as political candidates and has proven to provide fruitful analysis that would have not been possible without both the survey and digital data (Stier et al., 2019). For example, Quinlin et al. (2018) found the size of campaign budget was positively related to social media adoption by candidates and that the personality trait “openness” was positively associated with Twitter success. These findings would not have been possible without the linking of both social media 61 and self-reported data. In this study, linking social media and survey data provides for analysis of department and individual level impacts on social media outputs that is not possible without this method. Facebook posts themselves were also captured and used for a very limited qualitative exploration. The top ten interacted with posts and the top ten overperforming posts were captured to provide a small bit of context that is not captured through the measurement of other metrics. Overperforming scores were gathered from CrowdTangle. CrowdTangle computes an “Overperforming” score for each post in the data by first generating benchmarks from the last 100 posts from a given account of a given post type (link, image, video, etc.). The mean number of interactions is then computed from the middle 50% of posts at different ages (15 minutes old, 60 minutes old, 5 hours old, etc.), these are then set as the expected values. When a new post comes into the data from the account it is compared to the corresponding average. I examined each post, including video and pictures, and reading through a selection of the comments section. This represents only a glance into the interactions that are taking place on Facebook but was included to show the variety and depth of the platform. A much larger qualitative analysis would need to be conducted to fully understand the interactions on the platform. There are some inherent challenges of using social media data that should be addressed. It is important to note that the Facebook API is subject to rate limits. The amount of data that can be pulled in a timeframe is limited by each platform to a specific number of queries. I utilized CrowdTangle to ensure that the most data was captured, and limits are not exceeded. CrowdTangle allows for quick and easy access to public data for approved users. This allows for more data to be collected more quickly than would be possible with other methods. It is also important to recognize that social media companies are constantly updating and changing their 62 data limits and terms. Data was collected for multiple years, as CrowdTangle also allows for historic data to be collected. Detailed notes on the data collection process were taken, and there were no perceived effects caused by data limits themselves. There were a few accounts that had limited access to data due to their status. Thirteen total agencies could not have their social media data collected for various reasons. Four agencies completed the survey but refused the social media data collection. Three agencies who completed the survey did not have an active Facebook page at the time of the data pull. Another six agencies had pages with data that could not be accessed through CrowdTangle based on their status or page type. Data collection began with survey data, and shortly after, social media data collection began. The Facebook data was collected from January 2019 through January 1, 2023. For analysis of organizational theory variables in chapter 5, all four years of data were used but aggregated to the department level. This helps to control for history effects including major events that will be examined in chapter six. For chapter six, eight 2-month time periods were selected to examine Facebook use by departments before, during, and after certain events (see figure 3.1). The eight time periods include March-April 2019, 2020, 2021, and 2022, and May 25th – July 25th, 2019, 2020, 2021, and 2022. The first time periods served as a baseline point and included posts created during March through April of 2019, as well as May 25th Through July 25th of 2019. The next time period was March through April 2020, in order to capture data that occurred during the rise of COVID-19 in the US but before the death of George Floyd. The time period from May 25th through July 25th was selected to capture the murder of George Floyd and the following months to mirror the time that was captured for the COVID data. The first suspected community transmitted case of COVID-19 occurred around February 26th, 2020 and COVID-19 was officially declared a pandemic on March 11th, 2020 by the World 63 Health Organization (Timeline: How Coronavirus Got Started - ABC News, 2020). The third time period of Facebook data collected is from May 25th- July 25th, 2020. George Floyd was murdered on May 25th, so this data represents the time after George Floyd’s death and during subsequent protests. The final time periods (2021 and 2022) mirror the baseline data gathered and represent the follow up years. This study gets even more information about each time period through utilizing two-month blocks of data, yet a one-month timeframe has been used for similar studies of social media activity and engagement by government entities (Bonsón et al., 2017; Haro-de-Rosario et al., 2018). After all survey responses were collected, the social media data was merged into the database creating an agency level data set that includes the survey responses and the collected Facebook data. This dataset was then cleaned and prepared for analysis. Figure 3.1 Timeline of Facebook data for Chapter 6 After excluding participants who did not make it past twenty percent completion of the survey, there were 144 complete surveys to analyze. There was a total of 592 invitation emails sent, 278 emails confirmed opened, and twenty-four respondents that refused to take the survey or stated they felt survey was not relevant to their position. The 144 results represent a 24.32 percent response rate. There is reason to believe that PIOs were under intense pressure and limited for time due to the COVID-19 pandemic lasting effects, or the impact of other recent 64 events. The response rate also could have been increased by including more contact in a different format, such as phone calls but there was a lack of resources to complete phone calls to all potential participants as well as a lack of contact information including valid and direct phone numbers. Email contacts were easier to find and confirm for each PIO in the large sample. Examination of the PIOs who did not respond revealed no reason to believe they were different than those who did. The population served of those who did not respond was slightly lower than those who completed the survey (mean of 223,697 vs 279,374) and agencies had slightly more employees (mean of 634 vs 450 for those who finished). Region of those who responded, did not respond, or did not finish differed only slightly. Differences between the groups were not found to be statistically significant. Other Data sources There were a few other data points that were needed to address items from organizational theory. These variables were collected from the LEAR and 2020 Census data and attached to each PIOs response based on their agency or county served. To measure heterogeneity, I used the racial and ethnic diversity index from the 2020 census at the county or state (for state agencies) level. This index indicates how likely two people chosen at random will be from different race and ethnicity groups. The closer the index number is to zero, the lower the heterogeneity of that county, the closer to zero, the greater the heterogeneity (Rabe & Jensen, 2023). The LEAR data was attached by agency and includes agency characteristics collected from a few major data collection efforts including the 2008 and 2014 Census of State and Local Law Enforcement Agencies (CSLLEA). Crime rate was considered as a control but there was a lack of complete and reliable data for this variable. I attempted to calculate the crime rates from data pulled from the FBI Crime 65 Data Explorer (CDE) but found multiple issues. About 15% of agencies had no NIBRS reported from 2019 through 2022, another 12% only had data available for 2022 and only recently began reporting to NIBRS. Most of the state police data seemed to not populate in the CDE correctly as they have extremely low numbers, one state only showing reports for 2 or 3 violent crimes a year. Based on the lack of data and the lack of trust in the data that was available, I decided that there was not a reliable measure to use for crime rate for this study. Analytic Strategy First, descriptive analyses were conducted to answer questions about the role of PIOs in law enforcement organizations: Who are the personnel responsible for marketing police? What type of training do the personnel responsible receive? What are the current communication methods being used by police PIOs in the U.S.? What strategies are being used by PIOs to market police in media? What are the relationships between media and PIOs? What are the PIOs perceptions of portrayal of police by different media? How do PIOs utilize social media differently than traditional media? What strategies are being used to communicate with the public, do they vary based on format (social media vs. traditional media)? What changes has social media brought to police communication, specifically, what are the benefits and drawbacks? and What department resources (training, funding, time) are being put towards social media and how do PIOs collaborate with other units in the organization to share or collect information via social media? Multivariate analysis was then conducted to better understand the relationship between the independent organizational variables, individual level variables, and the dependent variables, outputs on social media (activity and engagement). The final research questions, “Do organization variables (size, funding, policy, leadership) impact the PIOs perceptions of the 66 value of social media for the department?, Do individual level PIO variables (understanding of the platform, training and experience with social media, motivation) impact activity and engagement of social media use? And are organizational or individual level variables better predictors of social media activity and engagement?” will be answered through multivariate regression analysis with organization and individual characteristics as independent variables and dependent variables of social media outputs. The collected data was used to test a model with both the characteristics of the police organizations and characteristics of the PIO impacting the department use of social media. The information gained from the descriptive analysis was used to further refine models before testing, ensuring that appropriate inputs are included. Based on organizational theory, it is expected that the findings of this study will reveal significant, but varied use of social media based on department level factors such as the budget put towards department use of social media, the size of the department, and the support of social media use by supervisors. These department level characteristics are expected to impact social media outputs along with individual PIO factors. The dependent variables are the social media outputs or department activity and engagement. As the two dependent variables capture the number of posts and the number of engagements with posts, these variables are continuous numeric. Regression analysis was adjusted to fit the distribution of the outcomes, using the most appropriate analysis for each. Variations explain differences in social media content, and the results highlight challenges and benefits of social media. In an area that is severely lacking in quality research, this study addresses multiple gaps in knowledge that will help to move law enforcement communication forward. These findings will improve police department understanding of how PIOs use social media and help PIOs, and agencies, better manage resources put towards the use 67 of social media. It will also identify responses to major events. These findings will be of great interest to law enforcement officials and other related practitioners. 68 CHAPTER 4: SURVEY RESULTS This chapter addresses the first group of research questions, updating our understanding of PIOs, as well as questions examining PIO management of social media. This chapter utilizes the survey data, as well as agency data from the LEAR data as this provided the most recent count of employees for the agency size. This chapter explores multiple issues related to research questions 1 through 6. The data generally shows that PIOs are responsible for police media and receive formal training in media. The data also indicate that PIOs use a mix of radio, newspaper, television, and social media but now heavily prefer social media for communication with the public. The data also generally indicates that relationships with news media organizations are positive and have remained so even with the more frequent use of social media. This chapter also presents data on PIOs perceptions of the biggest benefits and drawback of social media to their role. Updating Understanding of the PIO The demographic makeup of the PIOs completing the study is consistent with the findings from previous research (Motschall and Cao 2002). PIOs are largely white (68.8%) and male (59%) and have a mean age of 44.77 (see table 4.1). The majority of PIOs have served in their position 5 years or less (65%) and only about 10% have been in their position longer than 10 years. Many PIOs hold a college degree, with about 47.2% (n=68) holding a bachelor’s, 27.8% (n=40) with a master’s, and 1.4% (n=2) with a doctorate degree. Many PIOs also have had some sort of background or training in media. See table 4.1 for all PIO demographics. Table 4.1 PIO Demographics Variable N=144 Gender Male 85 % 59 69 Mean Std. Deviation 40 3 16 99 11 11 5 18 3 12 4 68 40 2 15 20 19 22 19 50 14 53 53 88 76 4 Table 4.1 (Cont’d) Female Prefer not to say Missing Age Race White Black Other Prefer Not to Say Missing Education High School Some college 2-year degree 4-year degree Master’s degree Doctorate Missing Rank Line Level (officer) Mid-Level Command Chief/Assistant Chief Civilian Missing Formal Training in media* Educational Background Occupational Background In-agency training Other None Satisfaction with Training Internal training External training Years in Current position 27.8 2.1 11.1 68.8 7.6 7.6 3.5 12.5 2.3 9.3 3.1 47.2 27.8 1.4 10.4 13.9 13.9 16.3 13.2 34.7 9.7 36.8 36.8 61.1 52.8 2.8 44.77 9.55 3.71 3.99 6.05 .92 .91 4.73 *Total will not add up to 100% as this item was select all appropriate responses 70 The demographics of PIOs are an important piece of the response to the first research question, “Who are the personnel responsible for marketing police?” The personnel that are directly responsible for marketing the police remain the PIO, as 95% of the agencies responding had specific staff dedicated to PIO duties, with about 72% full- time and 13% part-time PIO positions (see table 4.1). This is an increase in Sworn PIOs from the findings of Chermak and Weiss (2005) who found only 53.2% sworn PIOs in their sample. The staff size for PIO activities was also quite large. The majority of agencies responding had a PIO team that included about 2.8 total staff with the majority reporting 3 or less civilian staff (83.8%) or 5 or less sworn staff (79.2%)1.. As expected, the mean size of the agency (number of sworn and unsworn) significantly predicted the mean size of the PIO unit (R2 =.27, F (1, 129) = 14.45, p < .05). The size of the agency is based on data pulled from the LEAR dataset, specifically the Federal Bureau of Investigation (FBI) Police Employee data from 2014 as this was the most complete and recent data available for all responses. Chermak and Weiss (2005) found that PIO staff at the time consisted of a mean of 2.4 people, this study found that this has remained similar at an average of 2.83 staff. Chermak and Weiss (2005) also found that PIOs in the early 2000s had been in their current position for an average of 4.01 and that has increased slightly with respondents to this study reporting a mean of 6.05 years in their current role. The job of the PIO appears to be equal parts managing traditional media inquiries and promoting the department using social media strategies. The PIO position’s workload is made up of both traditional media management, with nearly 90% indicating PIO staff manage media inquiries, nearly 90% also indicate the PIO staff manage social media accounts. This is a 1 Outliers were removed for all analysis of PIO staff numbers as there seemed to be confusion on survey item. Outliers removed include those who reported PIO staff size greater than the total staff or within 30 staff different as the total number of staff used for calculations was dated. 71 noteworthy change in the role of the PIO over time and appears to be an important part of how police agencies now communicate with the public. Social media has been added to the traditional role of the PIO and seems to be a common communication method among agencies serving populations over 50,000 or employing at least 100 sworn officers. Table 4.2 Department Characteristics Variable N=144 % Mean Std. Dev. PIO status Full-Time Part-Time Less than part time I am not a PIO Missing Number of PIO staff* Sworn Civilian Total Who receives media training PIO staff Officers Command staff Chief Civilian staff Agency Size 150 or less 151-225 226-500 500+ Unknown** Missing Total Employees 103 19 7 2 13 134 29 81 47 17 35 36 34 35 3 1 71.53 13.19 4.86 1.39 9.03 93.1 20.1 43.8 32.6 11.8 24.3 25 23.6 24.3 2.1 .7 1.91 .89 2.83 3.01 .965 3.15 515.84 836.12 *This question was misunderstood by some resulting in abnormally high numbers reported. To adjust for this, I removed 14 outliers. I also used the most recent 2018 Census of State and Local Law Enforcement Agencies count for a total count of employees instead – available for 140 agencies. **Three agencies chose to fill out the survey but requested they remain completely anonymous so agency data could not be connected to their responses or social media data. 72 The duties of the PIO and their staff differ slightly from agency to agency, as could be expected, but have also grown to include social media and virtual tasks that were not previously part of the job. The most frequent task reported was communicating through social media with community groups with a mean of 12.57 times per month. The next most frequent task was the more traditional task of creating press releases for news media about 10 times per month. Table 4.3 shows the breakdown of each task frequency per month. Agency size was also positively and significantly related to press release frequency, press conference frequency, virtual press conference frequency, communicating with PIOs outside of the agency and press release frequency for social media. Agency size was not significantly related to frequency of attending community meetings, frequency of communicating with the public through social media, or frequency of contacting social media companies. 73 Table 4.3 Agency size and duty frequency in a typical month Variable n M SD 1 2 3 4 5 6 7 8 1. Total Employees 2. Press release 140 140 515.84 836.11 _ 9.86 9.13 .470** _ 119 2.07 4.60 .200* .317** _ 81 0.99 1.91 .429** .324** .258* _ 107 3.08 3.36 135 6.36 6.89 130 12.57 10.63 .084 .193* -.032 .154 _ .352** .515** .277** .501** .285** _ .144 .326** .034 .358** .110 .416** _ 123 9.44 9.46 .206* .472** .142 .438** .213* .266** .524** _ 69 0.84 3.80 -.003 .028 -.003 .018 -.057 .111 .309* .004 3. Press conference 4.Virtual press conference 5. Attend community meeting 6. Communicate with PIOs out of agency 7. Communicate through social media with community 8. Press release for social media 9. Contact social media company *p<.05 **p<.01 74 In examining PIOs’ collaboration within the organization, PIOs report meetings with groups within the agency most frequently and local news leadership least frequently. Many report not holding PIO staff only meetings which could be due to the small size of PIO staff or the meetings with other groups that were reported to happen daily by 20.1% (n=29) and at least weekly by 35.2% (n=51). PIOs meet with groups outside of the agency less often with 39.6% (n=57) reporting meeting less than once a month. Close to 85% of PIOs either did not contact local news media leadership or did so less than once a month, reflecting the earlier findings of Chermak and Weiss (2006). Note that PIOs were asked if they met with local news media leadership in a separate question, as this was a relationship of particular interest. This means that these news media leadership meetings could be also counted in the PIO staff meeting with groups outside of the agency. See table 4.4 for the complete breakdown of meeting frequencies. Table 4.4 Meeting Frequencies by type Meeting Type PIO Staff Only PIO staff and groups within agency PIO staff and groups outside agency Local News media leadership Daily Weekly Monthly Less Than Once a Month N/A Missing Total 21 (14.6%) 29 (20.1%) 7 (4.9%) 26 (18.1%) 57 (39.6%) 4 (2.8%) 144 (100%) 29 (20.1%) 51 (35.1%) 18 (12.5%) 23 (16%) 19 (13.2%) 4 (2.8%) 144 (100%) 5 (3.5%) 32 (22.2%) 22 (15.3%) 57 (39.6%) 23 (16%) 5 (3.5%) 144 (100%) 2 (1.4%) 4 (2.8%) 12 (8.3%) 63 (43.8%) 59 (41%) 4 (2.8%) 144 (100%) 75 Experience, Training, and Resources The next part of research question one, “What type of training doe the personnel responsible for marketing police receive?” was a bit more complicated. PIO personnel receive slightly varied training and come to their jobs with different experiences. There seems to be a variety of experiences that PIOs bring to the job with 23% (n=33) indicating an educational background (degree in media or public relations) and 34% (n=49) with a prior occupational background. Investment in the PIO position seems to vary with almost 33% of respondents indicating they received no formal training in the past year, but 61% (n=88) indicating some form of in-service training dedicated to media in the span of their career. The mean amount of general media training was reported to be about 18 hours per year but ranged from 1 hour to 81 hours per year. Chermak and Weiss (2005) had previously surveyed PIOs who stated they had an average of 48.8 hours of media training total. There was a wide variety of activities reported as formal media training including PIO specific classes, multiple conferences, FBI LEEDA courses, FBI academy courses, FEMA classes, outside vendor courses, and from local media. In general, PIOs are satisfied with both their training provided by the department (mean satisfaction of 3.71 out of 5) and training provided outside the department, but they are slightly more satisfied with external training (mean satisfaction of 3.99 out of 5). Only 4 (2.8%) PIOs reported no form of formal media training at all, leaving an overwhelming majority of 97% reporting some form of media training. This is slightly higher than Motschall and Cao’s (2002) prior findings of only 86.3% of respondents reporting that employees were trained in PIO techniques, but this could be due to differences in the type of question asked. Social media training was slightly different with 11.8% (17) reporting no formal training related to social media at all. The majority of PIOs have been using social media in their job for 76 less than 5 years (63.2%), but some have been using social media in their job more than 10 years (9.8%). On a personal level, PIOs have generally used social media much longer, with 59.4% reporting personal use for greater than 10 years and 30.6% (n=44) reporting personal use of more than once per day. Looking further at department resources dedicated to PIOs, the yearly budget designated specifically for PIO duties also varied from 0 (36%) to over $100,000 (8%). Considering resources devoted specifically to social media activities, only 37.5% (n=54) report dedicated funding, 57.6% (n=83) report dedicated staff, and 50.7% (n=73) report dedicated time. A few other resources were indicated such as IT staff and use of an outside vendor. Communication Methods Research question 2 asks “What are the current communication methods being used by police PIOs in the U.S. and what strategies are being used by PIOs?” Current communication methods vary based on preference and the issue being addressed, but social media was consistently reported as the most common method of communication for every issue asked about in the survey and radio was the least popular communication method for each issue (see table 4.5). This finding represents the clear shift to using social media to directly contact the community for a multitude of issues. Social media provides an efficient and direct mode of communication for the PIO, over which they have complete control. This immediate and complete control over the message that is being sent is likely why PIOs report such high favorability of social media as compared to more traditional forms. Even so, a few other modes, including television and newspapers remain popular for some of the basic duties of the PIO. Over 60% of PIOs reported the methods of TV and newspapers would be used for issued such as 77 requesting information from the public, reporting crime incident and arrest information, discussing new police programs, and responding to critical assessments. Table 4.5 Which media outlet would PIO be most likely to use by issue Method Request info from the public Build relationsh ips with the public Report crime incident info Issue Report crime arrest info Discuss new police programs Discuss new police policies Report departme nt awards and/or promotio ns 20% 41% 35% 96% Respond to critical assessmen ts 40% 64% 61% 84% 43% 64% 62% 98% 32% 45% 43% 97% 43% 68% 66% 89% 41% 73% 67% 84% 46% 62% 66% 94% 31% 56% 49% 76% Radio Newspa per TV Social media Other* * 3% 8% 7% 8% 8% 19% 7% 13% *Total will not equal 100% as multiple methods could be indicated for each issue type **Other media methods reported included department or county websites, in house apps, email distribution lists, and in person meetings Relationship with media and image This section addresses research question three, “What are the relationships between media and PIOs and what are the PIOs perceptions of portrayal of police by media?”. PIOs reported their relationship with news reporters to be fairly positive with 40% reporting an excellent relationship, and 50% reporting at least a good relationship (Average of 4.31 out of 5). This finding is comparable but slightly lower than that of Chermak and Weiss (2005) who found their respondents reporting a mean of 4.54. While this study did not capture the perspective of the media, prior research suggests that this relationship is likely symbiotic (Chermak & Weiss, 2006). News media and PIOs are still in fairly regular contact as PIOs reported an average of 2.52 calls from local news outlets per day. PIOs also perceive the quality of their agency’s image to be fairly positive as 32% reported their image as excellent, and 54% reported their image as 78 good (Average of 4.19 out of 5, see table 4.6). Agency size was significantly and negatively related to agency image in the local news, with smaller agencies indicating a better image. The quality of the relationship with local news reporters was not found to be significantly related to agency size but was positively, significantly related to the perception of agency image in local news, as would be expected. Table 4.6 Agency size and relationship with media and image n 140 142 M SD 515.84 836.11 4.31 .63 1 _ 2 3 -.006 _ 142 4.18 .67 -.190* .636** _ Variable 1. Agency Size 2. Quality of relationship with local news reporters 3. Agency image in local news *p<.05 **p<.01 PIO management of social media Research questions four and five, “How do PIOs utilize social media and how is this use different from traditional media? And what strategies are being used to communicate with the public and do they differ based on media type?” are explored in this section. It seems that social media has taken an important place in the PIOs job as many of their tasks are accomplished through social media and it is used for communication of multiple types of issues. Ninety-nine percent of PIOs stated that they use social media for their official duties. The most popular type of social media being used by PIOs is Facebook, but other types of social media have also become popular such as Twitter, Instagram, and YouTube (See Table 4.7). Another very popular 79 social media that was listed under the other option by 37 (25.7%) of PIOs was Nextdoor. Some of the other emerging or not as popular social media platforms listed include LinkedIn, Nixle, Neighbors (by Ring), TikTok, Flikr, and blog posts. Uses varied slightly by agency size with larger agencies tending to report more of a variety of types of social media use while agencies with 150 employees or less all reported using Facebook (see table 4.7). This could indicate that smaller agencies are still using Facebook, while larger agencies have been able to diversify their efforts to other platforms. Even so, most agencies report that Facebook was the social media platform used most often by their agency (n=102, 71%), so this was not related to agency size. Table 4.7 Social media types used by PIO for Public Information duties by agency size group Agency Size Type of social media Facebook Twitter (X) Instagram YouTube Other 150 or less 151- 225 226-500 500 or more Unkno wn* 35 (100%) 28 (80%) 21 (60%) 15 (43%) 9 (26%) 35 (97%) 31 (86%) 27 (75%) 23 (64%) 14 (39%) 33 (97%) 29 (85%) 22 (65%) 23 (68%) 7 (21%) 33 (94%) 34 (97%) 31 (89%) 27 (77%) 15 (43%) 3 (100%) 3 (100%) 3 (100%) 2 (67%) 1 (33%) Total Number n=144 139 125 104 90 46 Total Percentage 97% 87% 72% 63% 32% * Percentages within groups will not add to 100 as multiple social media options could be used and selected Social media has been part of the PIOs job for multiple years. About 41.67% (n=60) of PIOs state that their agency has maintained a social media presence for 10 years or more. Another 40.97% (n=59) reported a social media presence from 5-9 years and only 1.25% (n=18) reported a social media presence of 4 years or less. In general, PIO staff have regular authority to communicate using social media (n=137, 97%), but there are other staff who are sometimes 80 given access as well. The next most common role given routine access was the Chief (or other leadership) (n=53, 38%), followed by command staff (n=31, 21.5%), other civilian staff (n=26, 18.1%), and officers (n =12, 8.3%). These proportions remain similar when PIOs are asked who develops messages for social media, but with a slight drop in Chief to 28% (n=40) and Other civilian staff to 13% (n=19) (see Table 4.8). PIOs also noted that they often work with other units in their organization to release information to social media (n=127, 88.1%). Table 4.8 Who has access to department social media Access Type Chief PIO staff Role Command staff Officers Routine authority to communicate Develops messages 53 (38%) 40 (28%) 137 (97%) 136 (96%) 31 (22%) 23 (16%) 12 (9%) 12 (9%) Other Civilian staff 26 (18%) 19 (13%) Research question 6 asked “What changes has social media brought to police communication, specifically what are the benefits and drawbacks?”. Typically, social media is used to promote community events (82%) and report crime information (84%), but it is also often used to promote police programs and/or policies (63%), and awards or honors (62%) (see table 4.9). In reflecting on the benefits and drawbacks of social media use for their agency, the biggest benefit selected by PIOs was the connection with community members (94%), closely followed by the benefit of having a quick way to notify the public of safety concerns (91%) (see table 4.10). The biggest drawbacks cited by PIOs were the increased demand for quick information from the public (52%) and the time spent developing content for social media (36.1%). Notably, about 21% also wrote that some of the drawbacks to using social media 81 included misinformation shared by others/lack of control in comments (n=8), and “haters” or “trolls” who were attacking the department anonymously (n=11). The biggest barriers to social media use were identified as Responding to the public through social media (40.3%), measuring the impact of social media (31.9%), and lack of department resources devoted to social media (31.9%) (see table 4.10). One item that was listed as a drawback by some and a barrier by some PIOs was the public’s misuse of the social media page to report crime, especially when they should have called 911. Table 4.9 Typical Social media post Post type Number Percent Community Events Crime incident related information Police programs and/or policies Awards and honors Re-posts from other sources Other 113 116 87 85 49 20 78.5 80.6 60.4 59 34 13.9 *Percentage will not add to 100 as respondents could select multiple options Table 4.10 Benefits, Drawbacks, and Barriers of departmental social media use Benefit Number agreeing Percentage Connection with community members Quick way to notify public of safety concerns Public relations and reputation management Soliciting tips on crime from the public Recruitment 133 129 111 96 82 82 92.4 89.6 77.1 66.7 56.9 Table 4.10 (Cont’d) Intelligence gathering for criminal investigations Communicating with other government agencies Other Drawback Increased demand for quick information from public Time spent developing content for social media Loss of control of the meaning of a message Other Increased visibility of department Barriers Responding to the public through social media Measuring impact of social media Lack of department resources devoted to social media Adapting to new trends Training personnel to use social media effectively Using social media in a way that does not create a liability Designing policy for department social media use Gaining agency leadership support Building and managing IT infrastructure needed for social media Other 60 19 2 75 52 34 30 8 58 46 46 40 36 34 18 14 13 9 41.7 13.2 1.4 52.1 36.1 23.6 20.8 5.6 40.3 31.9 31.9 27.8 25 23.6 12.5 9.7 9 6.3 Social Media vs Traditional Media This section addresses the differences in social media and traditional media use in deeper exploration of research questions 4 and 5. PIOs seem to use social media similarly to traditional media but are more likely to choose social media for a variety of functions now that it is 83 available. While it seems that social media comes with its own set of benefits, drawbacks, and barriers, it is useful to know how PIOs use this medium when compared to more traditional forms of media. Strategies do seem to vary based on media type, as social media is seemingly more versatile. Social media is consistently identified as the best way to publicize each of the items identified, from community policing efforts to department policies and criminal investigations (see table 4.11). Social media seems to have been added onto PIOs workload, rather than completely replacing other media, as it also seems to now be the media of choice for many different purposes. It could be that social media is replacing some of the traditional media types such as radio and TV but there is a lack of prior published findings to directly compare to. Table 4.11 Best way to publicize, by situation Situation Newspaper Radio TV Media type Community policing efforts Successful criminal investigation (leads to arrest) Request for information about a case Department promotions/New hires Department policies 3 10 4 3 22 New initiatives 9 2 1 1 0 1 1 9 38 26 2 5 19 84 Social media 122 88 104 126 82 105 Other 3 2 3 5 20 3 CHAPTER 5: ORGANIZATIONAL THEORY FINDINGS This chapter addresses the following research questions: what department resources (training, funding, time) are being put towards social media and how do PIOs collaborate with other units in the organization to share or collect information via social media? Do organizational variables (e.g., size of the police organization, budget, extant policy, leadership; technological support/capacity) impact the PIOs’ perceptions of the value of social media? Do these variables impact an organization’s activity and engagement of social media use? Do individual-level PIO variables (understanding of the platform, training and experience with social media, motivation) impact activity and engagement of social media use? Overall, are organizational or individual- level variables better predictors of activity and engagement of social media use? PIOs’ perceptions of social media value. All analyses were conducted in SPSS version 29. Bivariate Analysis In this section, variable descriptives are laid out as well as bivariate analysis of individual, contingency, and institutional variables and their relationship to the outcome variables. First, I present the descriptives of all variables used in analysis. I find that the outcome variables are heavily right skewed, which indicates analysis for this chapter will need to be non- parametric. Table 5.1 includes descriptive statistics for all variables.2 First, when examining the two dependent variables, post counts for the specified time period (Jan. 1, 2019, to Jan 1, 2023) and 2 A few of these variables were recoded for clarity and analysis. The response “Not Sure” from items “Does management support department’s use of social media?”, “Has your agency sought outside sources of funding for social media?”, and “Is your agency CALEA accredited?”, were coded into “No”. These accounted for eight, four, and eighteen responses, respectively. Even after making these adjustments, it was clear that these variables lacked variance. It was determined that the variables measuring management support of social media and seeking 85 total interactions per post, data shows a lack of normal distribution. Agencies posted a mean of 1730 times, with a low of 105 and a high of 7990 posts. The average interaction per post was a mean of 293.3, with the minimum number of interactions being 37.95 and the maximum number of interactions being 1977.62 interactions. As is common in count variables, these variables are highly right-skewed. This is evident in the histograms for each (see figure 5.1). As these variables are not normally distributed, non-parametric tests will be utilized for analysis. Figure 5.1 Histograms of dependent variables external funding did not have enough variance to include in multivariate analysis with nearly 90% of respondents indicating their management supported their use of social media and nearly 90% reporting their agency did not seek outside funding to support their use. 86 Figure 5.1 (Cont’d) Table 5.1 Descriptive Statistics of all variables Variable Type Variable Label Dependent Independent Individual Number of Posts Average number of interactions per post Total number of media training hours in past year Years the PIO has used social media PIOs self-assessment of competence with social media 0 Completely incompetent 1 Somewhat incompetent 2 Neither competent nor incompetent 3 Somewhat competent 4 Extremely competent -99 Missing 87 n 131 131 Mean 1730.24 Std. Dev. 1420.56 % 293.2791 252.29 138 16.96 23.32 128 11.87 5.14 0 4 7 66 57 10 0.00% 2.78% 4.86% 45.83% 39.58% 6.94% Table 5.1 (Cont’d) Contingency PIOs rating of the importance of social media for their job 1 Not important at all 2 A little important 3 Moderately important 4 Very important 5 Extremely important -99 Missing PIOs rating of enjoyment of social media for job 1 Not at all 2 A little 3 A moderate amount 4 A lot 5 A great deal -99 Missing Social Media comfort index (enjoyment, importance, and competence) County Diversity Index 2020 Total number of employees Does Management support agency use of social media? 0 No or don't know 1 Yes -99 Missing Total PIO budget Total number of social media training hours in past year Tech Support available to PIO for social media 0 0 10 24 100 10 4 14 36 37 43 10 134 12.74 1.91 140 53.245 12.69 140 515.8357 836.11 7 129 8 102 25787.362 7 73874. 118 9.64 13.71 0 No 1 Yes -99 Missing 50 86 8 88 0.00% 0.00% 6.94% 16.67% 69.44% 6.94% 2.78% 9.72% 25.00% 25.69% 29.86% 6.94% 4.86% 89.58% 5.56% 34.72% 59.72% 5.56% Table 5.1 (Cont’d) Institutional Agency has sought external funding for social media 0 No 1 Yes -99 Missing 125 10 9 Agency is CALEA accredited 0 No 1 Yes -99 Missing PIOs types of formal training 0 1 2 3 4 -99 Missing PIO belongs to a professional organization 0 1 Frequency of contacting PIO from another agency 0 Never 1 Less than once a month 2 Monthly 3 Once a week or more -99 Missing PIO sought out other external resources for social media information 0 No external resources sought 1 Sought external resources Does PIO model their page after another LE agency 84 45 15 4 50 50 32 6 2 90 54 30 29 53 21 11 62 82 1 None at all 2 A little 24 43 89 86.81% 6.94% 6.25% 58.33% 31.25% 10.42% 2.78% 34.72% 34.72% 22.22% 4.17% 1.39% 62.50% 37.50% 20.80% 20.10% 36.80% 14.60% 7.60% 43.06% 56.94% 16.67% 29.86% Table 5.1 (Cont’d) 3 A moderate amount 4 A lot 5 A great deal -99 Missing How often PIO monitors other LE social media pages 1 None at all 2 A little 3 A moderate amount 4 A lot 5 A great deal -99 Missing Organizational Characteristics Region 1 New England 2 Middle Atlantic 3 East North Central 4 West North Central 5 South Atlantic 6 East South Central 7 West South Central 8 Mountain 9 Pacific -99 Missing Agency Type 1 Sheriff's office 2 Local police 3 Primary state police -99 Missing Total employees are greater than 100 0 No 1 Yes -99 Missing PIO manages social media 0 No 1 Yes -99 Missing 90 43 16 7 11 2 19 57 34 22 10 7 7 23 14 35 7 17 10 20 4 39 94 7 4 6 134 4 4 138 2 29.86% 11.11% 4.86% 7.64% 1.39% 13.19% 39.58% 23.61% 15.28% 6.94% 4.86% 4.86% 15.97% 9.72% 24.31% 4.86% 11.81% 6.94% 13.89% 2.78% 27.08% 65.28% 4.86% 2.78% 4.17% 93.06% 2.78% 2.78% 95.83% 1.39% The individual level variables indicate that social media is a critically important part of their job, and most are very prepared to use it. It is not surprising that the PIOs had a long history of personal social media use considering its cultural importance. On average, they have been engaged on social media for nearly twelve years. They received approximately seventeen hours of training on media-related issues. They indicated that they are somewhat or extremely competent in using social media for their job (85.41%), that it is extremely important for their job (69.44%), and they enjoy using social media a lot or a great deal (56%). These three variables were summed to form a media comfort index. Items were summed for this index because while they were only mildly statistically related, there was still a possibility of multicollinearity. Even so, summing allowed for each response to still maintain the same weight. An index was calculated to include in the future regression analysis because the items were related, measured on the same scale, and all represented individual perceptions of social media use for the department. The contingency variables attempt to capture how the organization’s support the PIO’s efforts in using social media, and several organizational-level characteristics are presented in Table 5.1. Overall, the PIOs are operating in diverse counties with a mean of 53.25, which is below the U.S. total score at 61.1. The diversity index ranges from 20.2 to 76.3, which also indicates variation in contexts worked. Note that there were agencies from all census regions, but a large number of respondents worked in the South Atlantic (24.31%) and the East North Central regions (15.97%). This was similar to the overall sample which was 21.4% South Atlantic and 13.8% East North Central. PIOs were representatives for all agency types as 27% were at a sheriff’s office, 65.28% worked at a local police agency, and 4.86% were employed by a state police agency. This breakdown is similar to national averages as the LEAR sample includes 19% 91 sheriff’s office, 80% local, and .3% state agencies. The slight overrepresentation of sheriff’s offices and state agencies is beneficial in this study to get a better understanding and to allow for some comparison. It is not surprising considering the sampling frame, but most agencies had at least 100 employees and almost all of them expected the PIOs to manage their social media accounts. Budget support varied considerably, ranging from no support to $600,000. The mean level of budget support was $25,787. Unsurprisingly, there was a wide range of budgets dedicated to the PIO and social media. Training for social media specifically varied as well with a mean of 9.64 hours per year and a maximum of over 60 hours per year. Sixty percent of the PIOs indicated that they had technological support, and ninety percent indicated that management supported their use of social media. The average number of employees was 516, with a range of 74 and 5771. In examining the institutional variables, sixty percent of agencies were not Commission on accreditation for Law Enforcement Agencies (CALEA) accredited (58.33%), and most PIOs did not belong to a professional organization (62.5%). Eighty-seven percent of the PIOs indicated that they did not pursue external funding. Almost fifty percent of the PIOs indicated that they did not model their social media pages from other departments, but the vast majority indicated that they view social media pages from other agencies (approximately 75%). A majority of PIOs reported seeking external sources for information about social media use (56.94%). PIOs also reported a variety of types of formal training in media communication including a degree in a related field, prior occupation in media, in-service training, and external training such as the Law Enforcement Executive Development Association (FBI-LEEDA) and 92 Federal Emergency Management Agency (FEMA) courses, with sixty-one percent reporting at least two of these forms of training. Below are the hypotheses which were discussed in greater detail in chapter 2: Individual level hypotheses: H1: The amount of all media training will be positively related to both department activity and citizen engagement. H2: The length of time an individual has used Facebook will be positively related to both department activity and citizen engagement. H3: PIOs who feel competent in their use of social media for the department will be positively related to both department activity and citizen engagement on Facebook. H4: PIOs who feel Facebook is important for their department will be positively related to both department activity and citizen engagement on Facebook. H5: PIOs who enjoy using Facebook for their department will be positively related to both department activity and citizen engagement on Facebook. Organizational Level Hypotheses: Contingency: H6: Heterogeneity of the community served will be positively related to the department’s use of social media. H7: Size of the agency will be positively related to both department activity and citizen engagement on Facebook. H8: Command Staff support will be positively related to both department activity and citizen engagement on Facebook. 93 H9: Amount of department resources (budget, time spent on social media training, technological support) devoted to social media will be positively related to department activity on Facebook. Institutional: Isomorphism, Coercive: H10: Agencies who have received external funding for departmental social media activities will be more likely to use social media more frequently. Normative: H11: Accredited agencies will use social media more frequently. H12: PIOs with more formal training will use social media more frequently and have greater citizen engagement. H13: PIOs who belong to professional organizations will use social media more frequently and have greater citizen engagement. H14: PIOs who seek outside sources of information on how to utilize social media will use social media more frequently and have greater citizen engagement. Mimetic: H15: PIOs who regularly monitor and/or model their departmental social media use after other law enforcement agencies will use social media more and will receive greater engagement. Spearman’s correlations were used to explore the bivariate relationships between the continuous and ordinal predictors (Table 5.2) and Mann-Whitney U tests were used for dichotomous variables (Table 5.3, 5.4, 5.5). Continuous Predictors Table 5.2 presents correlation results for all continuous and ordinal variables. The non- parametric Spearman’s correlation was used to assess the strength and direction of relationships. 94 This test was chosen because it allows for ordinal independent variables, is not sensitive to outliers, and does not require normally distributed variables. Overall, there is not strong support for any of the hypotheses, moderate support for two individual-level variables (H3-competant, H5-enjoy), and one contingency variable (H7-size). There was weak support for one additional individual-level variable ((H2-length of time), and two contingency variables (H6-heterogenity; H9-department resources). None of the institutional variables indicate a relationship with the two dependent measures. In highlighting the results pertaining to the individual level hypotheses, the amount of time a PIO used social media was positively, but weakly, related to the number of posts and not the post interaction rate (H2). The PIOs perception of the importance of social media to their job was also positively and mildly related to the number of posts and not related to the post- interaction rate (H4). Competence and enjoyment of social media use were both positively, moderately related to post frequency and not related to interaction rate (H3 and H5). Hypothesis 1 is not supported as the overall amount of media training was not related to post frequency or post interaction. Turning to the contingency variables, heterogeneity of the county was not related to post frequency but was negatively, weakly related to the interaction rate with posts (H6). The agency size was strongly and positively related to the post frequency while also being positively but weakly related to the interaction rate (H7). Command staff support did not differ enough to be included in the analysis (H8). The budget spent on PIOs was weakly and positively related to post frequency but not related to interaction rate (H9). The final six hypotheses pertaining to institutional variables did not receive support (H10-H15). The variables, formal training, contacting other PIOs as a resource, seeking external 95 sources for social media information, modeling other agencies’ social media, and monitoring other agencies’ social media were all found to have no relationship with the number of posts or the interactions per post. 96 Table 5.2 Correlations of all continuous and ordinal variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1. Posts number 2. Interactions per post 0.144 3. Total media training hours 0.075 0.034 4. Years using social media .212* 0.033 -0.093 5. Competence in social media .330** 0.126 .259** .382** 6. Importance of social media .257** 0.136 0.156 .271** .264** 7. Enjoyment of social media .302** 0.103 .234** 0.167 .447** .493** 8. 2020 diversity index 9. Agency size - total employees 10. PIO budget 11. Total social media training 12. Types of social media training 13. Contacts others as resource 0.162 -.250** 0.039 0.005 0.118 0.139 0.157 .416** .247** 0.083 0.164 .274** 0.161 .217* .221** .288** 0.023 0.16 .257* .295** -0.015 0.132 0.097 .270** 0.105 0.039 .711** -0.1 .195* 0.149 .238** 0.159 0.152 0.204 .176* 0.029 -0.015 .263** 0.151 0.024 0.069 .168* .255** .309** -0.049 0.091 -0.121 .252** 0.033 -0.012 0.14 0.144 0.139 .188* 0.185 .403** 0.126 0.03 0.085 0.007 -0.027 14. PIO seeks external sources for social media 15. PIO models other LE social media 16. PIO monitors other LE social media * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2- tailed). -0.059 -0.159 0.099 0.141 -0.027 0.06 0.141 .174* 0.133 0.149 0.161 0.027 0.118 0.133 0.115 -0.077 -0.114 0.106 0.025 0.088 0.059 -0.163 -0.028 -0.003 .209* 0.113 -0.128 0.094 0.118 0.09 0.113 0.036 -0.036 .204* 0.023 .248** 0.16 .375** 97 Dichotomous Variables Table 5.3 presents the means of dependent variables by the dichotomous predictor variables. While these will be tested with a non-parametric, Mann-Whitney U test, it was helpful to first examine the overall differences in means. Visually, it seems that these groups may differ slightly in their social media activity practically but not in a significant way (See table 5.3). To confirm, nonparametric tests were employed to compare the three dichotomous, independent variables to the dependent variables. There were no significant differences in post number of interaction rate. Table 5.3 Mean of each dependent variable by binary group Mean Posts Mean Interaction rate Tech Support is available for social media Agency is CALEA Certified No Yes 1760.42 1762.95 344.83 272.34 No Yes 1671.64 1840.90 295.22 241.71 Member of any professional association No Yes 1583.84 1975.22 275.83 322.48 Separate Mann-Whitney U tests were performed to evaluate whether mean number of posts and mean rate of interaction differed by tech support availability, CALEA certification, or membership in a professional association. The results (see tables 5.4; 5.5) indicated that there 98 was no significant relationship between tech support, being CALEA certified, and being part of a professional organization with the number of posts and the mean rate of interaction. Based on these outcomes, there is no support for H9, H11, and H13. Table 5.4 Post count means and Mann-Whitney U Measure Response No Yes No Yes No Yes Tech Support CALEA Professional association Table 5.5 N 48 76 76 42 82 49 Mean rank Mann- Whitney U p-value 63.49 61.88 58.92 60.55 62.02 72.65 1776.5 0.807 1640 0.805 2335 0.121 Interaction rate means and Mann-Whitney U Measure Response Tech Support CALEA Professional association No Yes No Yes No Yes Mean rank Mann- Whitney U p-value 68.31 58.83 58.07 62.10 63.78 69.71 1545 0.152 1705 0.54 2191 0.387 N 48 76 76 42 82 49 99 Multivariate analysis Multivariate analyses were used to answer the tenth research question: “Are organizational or individual level variables better predictors of PIO social media activity and engagement?” In addition, although it appears from the bivariate analyses that most of the predictors did not strongly impact the outcome variables, these analyses will also show the relative influence of the variables. Because the dependent variables are count data and follow a non-normal distribution that is skewed, Poisson and negative binomial regression were considered for these analyses. The variances of the dependent variables were much larger than the means, and the dispersion of the data was too high for Poisson models, thus negative binomial regression was determined to be most appropriate to handle the overdispersion of the data (Berk & MacDonald, 2008). There were no artificial or true zeros, so a zero-inflated model was deemed inappropriate. Table 5.6 presents the results from the multivariate analyses. Three negative binomial analyses were conducted for the two outcome variables: a model for each dependent variable that included only individual, contingency, and institutional variables. Type of agency was used as a control in all models and a final model was produced that included only the significant variables from these analyses3. The first model examined the individual independent variables including training hours in media, the number of years using social media as an individual, and an index combining self- rated competence, view of importance, and enjoyment of social media. The final three were combined into a single factor as these were found to be significantly correlated with one another 3 Region was also considered for use as a control in the analysis. Ultimately, due low sample size, in order to keep the variable amount lower and power of statistical analysis higher it was not included. Models were ran including both region and agency size and no substantial differences were indicated in the outcomes compared to the findings presented here. Future full tests of the theory should include region when possible. 100 (see table 5.2). This was a positive and significant (b = 0.143, SE = .0357, p<.001) predictor of the number of posts. The incidence rate ratio (IRR) indicates that during the time frame of the study, PIOs with greater competence, enjoyment, and view of the importance of social media were predicted to post 1.15 times that of those with lower competence, enjoyment, and view of the importance of social media (holding the remaining predictors constant). In other words, each unit increase in the competence, enjoyment, and perceived importance of social media resulted in a 15% increase in number of posts by that PIO. The other predictors did not yield significant results (see Table 5.6), and the model with the interaction rate outcome variable yielded no significant results. The next model focused on contingency variables and included the following independent variables: heterogeneity (diversity index), size of agency, and the number of resources (social media training hours and technical support for social media). Some variables, such as budget and social media training hours were significantly correlated and while the strength was weak, the budget variable had many missing cases. The budget variable was excluded from these analyses, and so was management support because of limited variability. Diversity index (b=0.013 SE = 0.341, p<.01), number of total employees (b=0.00, SE = 0.00, p< 0.001), and the hours spent on training (b = 0.026, SE = 0.013, p<0.05) were all significant predictors of the number of posts. The diversity index (b=-0.016, SE = 0.005, p<.001) and the number of employees (b=0.00, SE = 0.00, p< 0.001),) were also significant predictors of interaction rate but in this case, the diversity index was negatively predictive (See table 5.6). 1s indicate that during the time frame of the study, each unit increase in diversity resulted in 1.3% increase in number of posts and a 1.6% decrease in interactions (holding the remaining 101 predictors constant). Practically, these differences may not be very large, but the relationship is worth exploring in future research with more robust measures. The next models focused on the following institutional independent variables: CALEA accreditation, formal media training, professional organization membership, information seeking, and PIO mimetic behavior including monitoring and modeling after other law enforcement social media. This model performed very poorly, further showing that the institutional level variables in this study were not good predictors of social media activity (see table 5.6). The final model presented in Table 5.6 included the significant variables from the contingency and individual models. The individual variable included in this analysis was the index measuring the enjoyment, competence, and importance of social media to the PIO. The contingency variables included the diversity index, the total employees, and the training hours in media. The agency type was included in the model as a control. All variables except the number of training hours remained significant in this model (see table 5.6). The PIO index of competence, enjoyment, and importance of social media (b=0.095, SE= 0.034, p<0.01), the diversity index (b=0.01, SE=0.005, p<0.05), and total employees (b=0.00, SE=0.00, p<0.01) remained significant in this combined model. 102 Table 5.6 Negative binomial outcomes Variables Individual Training Hours Time using social media PIO Social Media use Index Agency Type Contingency Diversity Index Total Employees Training Hours TechSup Agency Type Institutional CALEA Training types Professional Organizations Contacts other PIOs Seeks External Information Model’s others Monitors others Agency Type Contingency and Individual PIO Social Media use Index Diversity Index Total Employees Training Hours Agency Type B -0.002 0.02 0.143 0.04 0.013 0 0.026 0.017 -0.014 0.076 0.079 0.22 -0.015 -0.048 -0.035 0.103 0.045 0.095 0.01 0 0.017 0.011 Post Frequency IRR 0.998 1.02 1.153 1.041 se 0.003 0.014 0.036 0.057 0.341 0 0.013 0.126 0.057 0.143 0.079 0.149 0.075 0.153 0.067 0.071 0.06 0.0339 0.0049 0 0.0127 0.0567 1.013 1 1.026 1.017 0.986 1.079 1.104 1.246 0.985 0.953 0.965 1.109 1.046 1.1 1.01 1 1.018 1.011 103 Sig 0.416 0.145 <.001 0.487 0.01 0.001 0.044 0.896 0.801 0.596 0.21 0.14 0.839 0.754 0.598 0.149 0.446 0.005 0.035 0.003 0.169 0.85 Interaction Rate IRR 0.999 1.011 1.059 1.048 se 0.497 0.003 0.015 0.038 0.005 0 0.013 0.12 0.054 0.131 0.069 0.137 0.07 0.142 0.064 0.067 0.051 0.984 1 1.019 0.866 0.939 1.09 1.126 1.149 0.914 1.012 0.957 0.913 1.07 b -0.001 0.011 0.057 0.047 -0.016 0 0.019 -0.144 -0.063 0.087 0.119 0.139 -0.09 0.012 -0.044 -0.092 0.068 Sig 0.733 0.486 0.133 0.369 <.001 <.001 0.14 0.232 0.241 0.51 0.087 0.313 0.198 0.934 0.496 0.174 0.188 Post Content While statistical analysis to this point has illuminated some of the activities of PIOs on social media, there is nuance that this quantitative data cannot capture. To begin to address this, a small sample of cases were drawn from some of the top performing and top interacted with posts during the sample period. I examined each post by reading through the post, watching video, or viewing images, and reading through a selection of comments including the “most recent”, and the “most relevant” filtered on Facebook which included those comments that had the most interactions. This represents only a glance into the interactions that are taking place but serves to show the variety occurring in this digital space. The themes and general characteristics are explored briefly in this section. First the top ten total interactions posts were examined to understand possible themes that may encourage interaction. Of these top ten interacted with posts, eight posts were original or native video posted to the agency page. Interestingly, the top interacted with post is from March 24,2020 and is still (as of 2/14/2024) getting new comments and shares. Half of the top ten were posted in 2019, the other 5 were spread out from January 2020 through May of 2021. Two agencies had two posts in the top ten. Overall, there were no consistent themes or content type, but these posts actually represented the variety of content that is commonly posted on police social media pages. There were two posts asking for tips but, interestingly, many of the interactions were not related to spreading the post but rather commenting on racial tension, the high quality of a security camera, or the creativity of the suspect. Three of the posts involved interactions with or rescue of a child. Three posts involved humor, with two being created specifically for social media by the department to be used as a humorous post, one was specifically a dance challenge response. One video demonstrated new technology in the form of 104 less than lethal options now available, posted in 2019. Two posts had multiple comments and interactions occurring between viewers in the comments surrounding racial tension and these were both posted in August and September of 2020, after the murder of George Floyd. Instead of focusing only on the interaction rate, overperforming posts were also considered to understand why an agency’s post might perform particularly better than their other posts. CrowdTangle assigns the overperforming score to each post based on past post performance for the individual page. Based on how this is calculated, the top overperforming posts represent the biggest differences in the account’s normal activity, not necessarily the top performing posts in the dataset overall. Eight of the top ten overperforming posts included native video and the remaining two included an image. Three posts were posted in 2019, three in 2020, two in 2021 and one in 2022. Six of these posts, including the top three, were also posts that were in the top ten interacted with posts. Of the six that were also from the top interacted with: two included helping children, one was humor created specifically for social media, and two asked for tips but also included racial tension. Of the posts that were not in the top ten interacted with: two were retirements, one was K-9 related, and one was based on humor and a dance challenge. 105 CHAPTER 6: HIGH PROFILE CASES This chapter will address the final research questions regarding PIOs and high-profile cases. • How did COVID-19 impact PIO strategies and use of social media? • How did the death of George Floyd impact PIO strategies and use of social media? • How did each event impact social media engagement? COVID-19 will be discussed first with analysis of respondent’s thoughts on how the pandemic impacted their job as a PIO and use of social media. These responses are analyzed by agency size groups to better understand if and how the agency size may have impacted responses to these high-profile cases. The second portion of the chapter will be dedicated to analyzing perceptions of how the murder of George Floyd and the subsequent protests impacted the PIOs’ job and use of social media. Analysis will also include brief examination of social media engagement and posting frequency before and after each event. COVID-19 Responding PIOs largely (71% responding “definitely”) agreed that their jobs were impacted by the COVID-19 pandemic. This response was not related to the size of the agency and was fairly consistent across groups by agency size (69-74%). PIOs also stated what specifically changed about their job. Fifty-two percent mentioned the movement of in person events to online venues was inconvenient can caused additional work (see table 6.1 for full breakdown by agency size). Some also stated that pandemic hindered their ability to conduct community outreach as this type of interaction had to be moved to a new medium. The other top topics discussed were the change in the type of information produced (28%) and the increase in 106 information put out through social media (26%). A couple PIOs specifically noted the change in the medium causing issues such as increase in demand for native video (they needed to create rather than relying on reporters), or the change to a live format creating an environment where videos could not be cut or edited by a reporter meaning, “fewer missteps or stutters would be forgiven.” Table 6.1 Was your job as a PIO impacted by COVID-19? By agency size Impacted Definitely yes Probably yes Might or might not Probably not Definitely not Total 150 or less 22 (68.8%) 5 (15.6%) 1 (3.1%) 151-225 23 (71.9%) 3 (9.4%) 1 (3.1%) Agency Size 226-500 22 (68.8%) 2 (6.3%) 1 (3.1%) 500+ 25 (73.5%) 3 (8.8%) 1 (2.9%) Total 92 (70.8%) 13 (10%) 4 (3.1%) 2 (6.3%) 2 (6.3%) 32(100%) 2 (6.3%) 3 (9.4%) 32(100%) 4 (12.5%) 3 (9.4%) 32(100%) 4 (11.8%) 1 (2.9%) 34(100%) 12 (9.2%) 9 (6.9%) 130(100%) On the other hand, PIOs did not believe that COVID-19 changed their relationship with the news media with 82% responding it did not change. This did not differ significantly by agency size. Only about 2% of responding agencies reported a worse relationship with news media and about 12% reported a better relationship. Table 6.2 shows the complete breakdown by agency size. Table 6.2 Did your relationship with news media change as a result of COVID-19? By agency size group Changed relationship? Yes, it became worse Yes, it became better 150 or less 151-225 Agency Size 226-500 500+ Total 2 (6.1%) 0 1 (3.1%) 0 3 (2.3%) 1 (3.0%) 3 (9.4%) 6 (18.8%) 5 (15.2%) 15 (11.5%) 107 Table 6.2 (Cont’d) No, it did not change Not sure Total 30 (90.9%) 29 (90.6%) 22 (68.8%) 25 (75.8%) 0 33 0 32 3 (9.4%) 32 3 (9.1%) 33 106 (81.5%) 6 (4.6%) 130 The majority of PIOs reported that their use of social media did not change as a result of COVID-19 (55%), and about 39 percent reported using it more (see table 6.3). Of those who did report using it differently, 88 percent stated they posted more frequently (34% overall), and 78 percent stated they posted more public health information (30% of total respondents). Most PIOs who stated their use changed (60%) also reported holding more virtual events through social media (see table 6.4). Table 6.3 Did your use of social media change as a result of COVID-19? By agency size group 150 or less 151-225 Agency Size 226-500 500+ Total 11 (33.3%) 11 (34.4%) 12 (37.5%) 16 (48.5%) 50 (38.5%) 0 3 (9.4%) 1 (3.1%) 1 (3.1%) 5 (3.8%) 21 (63.6%) 18 (56.3%) 17 (53.1%) 15 (45.5%) 71 (54.6%) 1 (3.0%) 33 0 32 2 (6.3%) 32 1 (3.0%) 33 4 (3.1%) 130 SM use change? Yes, we used it more Yes, we used it less No, out use did not change Not sure Total Table 6.4 How did your social media use change as a result of COVID-19? (N = 50) Activity Posted more frequently Number Responded more frequently to public’s questions Held more virtual events Percent 88% 42% 60% 44 21 30 108 Table 6.4 (Cont’d) Posted more public health information George Floyd 39 78% It appears that the murder of George Floyd did impact agencies of different sizes somewhat, but again no significant differences were found. The majority of PIOs responded that the murder of George Floyd definitely impacted their job as a PIO (52%), and another 14.2% responded that it probably did (see table 6.5). These PIOs were then asked to briefly describe how their job was impacted. The most discussed topics include attacks directed at the police or anti-police rhetoric (31%), increase in public interest in policies (27%), increased communication or interaction with community (22%), and increase in effort to combat the anti- police rhetoric (17%). One PIO also explicitly mentioned a shift in tone stating, “we found that our content needed to be straightforward and not humorous”. Another mentioned that along with the public being more critical of positive posts, officers also were more reluctant to have their names and pictures published on the agency social media pages. Table 6.5 Was your job as a PIO impacted by the murder of George Floyd? by agency size Impacted Definitely yes Probably yes Might or might not Probably not Definitely not Total 150 or less 10 (32.3%) 7 (22.6%) 7 (22.6%) 151-225 18 (56.3%) 3 (9.4%) 4 (12.5%) Agency Size 226-500 14 (43.8%) 4 (12.5%) 4 (12.5%) 500+ 24 (75.0%) 4 (12.5%) 2 (6.3%) 5 (16.1%) 2 (6.5%) 31 4 (12.5%) 3 (9.4%) 32 7 (21.9%) 3 (9.4%) 32 1 (3.1%) 1 (3.1%) 32 Total 66 (52.0%) 18 (14.2%) 17 (13.4%) 17 (13.4%) 9 (7.1%) 127 The majority of PIOs reported that their relationship with news media did not change after the murder of George Floyd (78%) (see table 6.6). PIOs also mostly reported that their use 109 of social media did not change after the murder of George Floyd as only about twenty percent reporting using social media more after (see table 6.7). Table 6.6 Did your relationship with news media change as a result of the murder of George Floyd? by agency size group 150 or less 151-225 Agency Size 226-500 500+ Total 0 5 (15.6%) 2 (6.3%) 3 (9.4%) 10 (7.9%) 2 (6.7%) 4 (12.5%) 5 (15.6%) 1 (3.1%) 12 (9.5%) 28 (93.3%) 22 (68.8%) 22 (68.8%) 26 (81.3%) 98 (77.8%) 0 30 1 (3.1%) 32 3 (9.4%) 32 2 (6.3%) 32 6 (4.8%) 126 Changed relationship? Yes, it became worse Yes, it became better No, it did not change Not sure Total Table 6.7 Did your use of social media change as a result of the murders of George Floyd? By agency size group 150 or less 151-225 Agency Size 226-500 500+ Total 5 (16.1%) 4 (12.5%) 5 (15.6%) 12 (37.5%) 26 (20.5%) 1 (3.2%) 4 (12.5%) 4 (12.5%) 1 (3.1%) 10 (7.9%) 25 (80.6%) 24 (75.0%) 22 (68.8%) 18 (56.3%) 89 (70.1%) 0 31 0 32 1 (3.1%) 32 1 (3.1%) 32 2 (1.6%) 127 SM use change? Yes, we used it more Yes, we used it less No, out use did not change Not sure Total Table 6.8 How did your social media use change as a result of the murder of George Floyd? (N =26) Activity Posted more frequently Number 19 Percent 73.1% 110 Table 6.8 (Cont’d) Responded more frequently to public’s questions Held more virtual events Other 14 4 5 53.8% 15.4% 19.2% Of those who did report their social media use changed, most (73%) of the PIOs reported posting more frequently after the murder of George Floyd and many (54%) reported responding more frequently to public questions (see table 6.8). The departments reporting “other” here all commented on increasing trust and transparency through their social media in some way. Some suggested that they live streamed video more and featured unedited bodycam footage on their page. Some PIOs created more avenues for community feedback, and one suggested they added more information in order to “humanize the badge.” While there were much fewer agencies reporting negative impacts of the murder of George Floyd on their social media use, the ones who were impacted reported some specific difficulties. One agency reported shutting their whole Facebook page down. Based on Facebook data, it appears that they kept their page dormant until coming back mid-December of 2020. Another PIO reported posting less frequently due to negative views of the department and then increasing “back to normal” as things “calmed down” in the community. Impact on social media outputs To examine how these events may have impacted the social media outputs used as dependent variables in chapter 5, additional analyses were completed. As we know, these major events occurred fairly close together and there is likely overlap in effects. COVID-19 persisted into when George Floyd was murdered. Interestingly, the chart of post frequency does seem to 111 show a difference in the frequency of posting but it had started to decrease before the death of George Floyd. In the year over year analysis, there is a disruption in the frequency of posts during both COVID-19 and after the George Floyd murder. In 2020, posts made from March 1st through April 30th made up about 20% of all posts made that year as compared to an average of around 16% the year prior and years after, this could be due to the decrease from May 25th through July 25th that was also observed (see table 6.9). In this period, after the murder of George Floyd, during 2020 the frequency of posts made up only about 15% of the year total for all agencies as compared to about 17% the year prior and two years after. Further, there were different percent changes observed when compared to 2020. For the COVID-19 period, in the year following, there were 18.21% fewer posts made during that period (see table 6.10). All indicating an increase in frequency of posts during COVID-19, as was suggested by the PIOs responses. Similarly, but in the opposite direction, there was a 13.74 percentage increase in 2021 when comparing the time period directly following George Floyd’s murder, even though many PIOs reported that their use did not change and only 7.9% suggested using it less during 2020. Table 6.9 Year over Year change in post frequency from COVID-19 in 2020 (March 1 through April 30th) Year Number of % of total Difference % change posts posts that from 2020 from 2020 year 2019 11225 16.7% -1162 -9.38% 112 Table 6.9 (Cont’d) 2020 12387 20.4% - - 2021 10131 16.4% -2256 -18.21% 2022 10914 16.6% -1473 -11.89% Table 6.10 Year over Year change in post frequency from George Floyd Murder in 2020 (May 25th through July 25th) Year Number of % of total Difference % change posts posts that from 2020 from 2020 year 2019 11423 17.0% 2076 22.21% 2020 2021 9347 15.4% 0 10631 17.2% 1284 13.74% 2022 11219 17.1% 1872 20.03% 113 CHAPTER 7: CONCLUSION Police PIOs must work in and respond to a media landscape that is in constant flux, with new technologies being introduced often. Social media represents just one new media that has altered the way society communicates. This new format allows police more direct lines of communication with the broader public. This dissertation pushes forward research on PIOs by examining three key things: updating understanding of the role of the PIO, exploring the application of organizational theory to understand PIO use of social media, and capturing PIO social media response to two high-profile events. In summary, Chapters 1 and 2 highlighted the importance of this study to fill in the gaps of our knowledge about the current role of the PIO, what individual, contingency, and institutional variables might impact the use of social media, and understanding how current major events may impact PIOs use of social media. Organizational theory is presented, including contingency and institutional theory, and elements of each theory are discussed as potential influences on PIO use of social media. Chapter 3 describes the methods used in this study. Using prior studies and theory as a framework, this dissertation sought to be one of the first to gather information about social media use from PIOs themselves through a national survey and then to connect these responses to social media outputs. Based on when the data was gathered for this study and the occurrence of two tragic events, it was also positioned to capture PIO response to COVID-19 and the murder of George Floyd. Chapter 4 started with updating the understanding of PIOs and found that the demographic makeup and average amount of PIOs in an agency were fairly similar to the findings of Chermak and Weiss in 2005. Differences, as compared to findings from Chermak and Weiss (2005), were found in the amount of sworn PIOs (increased) and workload. The increase 114 in sworn PIOs could be due to an increase in professionalization of the role and the opportunities that social media present. The workload of the PIO included management of traditional media but now also includes managing social media accounts. In fact, the most frequent task reported was communicating through social media with community groups, overtaking the more traditional task of creating press releases. PIOs reported a variety of different media training activities, with only slightly less reporting having no formal social media training. There was a wider range observed in the experience in use of social media though, as some PIOs have been using it in their job twice as long as others. This study took place at an ideal time to capture the change in roles based on social media because most respondents reported their agency had used social media for 5 years or more. This is likely long enough for it to become part of the PIOs role, which was confirmed by the high numbers of PIOs reporting preferring social media for certain job tasks. In general, it was found that PIOs now prefer social media as a mode of communication for multiple job tasks, with the preferred social media still being Facebook. Larger agencies were found to utilize more of a variety of platforms as compared to smaller ones. PIOs still report high favorability of traditional media (such as TV or newspaper) for some specific job tasks, but social media now outranks these media. Another interesting finding that had not been captured from PIOs specifically until this point was the report of benefits, drawbacks, and barriers of social media to their role. The biggest benefit was connecting with the community, but on the other hand, the biggest drawback was increased demand for quick information from the public. It seems that social media removed the barrier of traditional media gatekeepers who previously dealt with information demand and now PIOs are seeing that demand firsthand. In general, social media seems to be something that 115 has transformed the PIOs role, is favored for multiple uses by PIOs, and has become a normalized part of the job. Chapter 5 demonstrated that there may be particular individual and organizational factors that influence the actual outputs and interactions seen on law enforcement social media pages. While this was more of an exploratory test of theory, several theoretical perspectives informed the items that were included, and this study lays the groundwork for future study by noting what relationships may be connected to measurable social media outputs. The strongest evidence was found to suggest that individual PIO’s perception of their competence with social media, their enjoyment of social media, and their view of importance of social media to their job positively influences the number of posts they produce but not the interactions that the public has with these posts. In support of elements of contingency theory, diversity of the area they serve and the total employees in an agency were positively related to frequency and while total employees were also positively related to interaction rate, diversity was actually negatively related to interaction rate. To further capture the nuances that may be missed by these simple social media outcomes, this study examined some posts that were the top interacted with during the study period. Through brief qualitative analysis, I found there were no clear themes in these posts, but most included original video. This highlights the fact that social media is inherently (and perhaps preferably) multi-modal, allowing for original video to be paired with original text, all curated by the PIO. Interaction also seemed to be linked to recent events, with more racial tension observed in comments during the months following the murder of George Floyd, this may have influenced the finding that areas with greater diversity had greater number of posts, but lower interaction rates. Also interestingly, some of the top ten most interreacted with posts were also on the top list 116 of overperforming posts. This means that some of the most interacted with posts are not “normal” for the agencies posting them but rather represent exceptions. Chapter 6 dove deeper into the understanding of the impact of high-profile events on the job of the PIO and their use of social media. It was revealed that COVID-19, unsurprisingly, moved more of the PIOs job online and slightly increased the posts on law enforcement pages. PIOs also reported a slight change in content and an increase in need to respond to the public through social media. PIOs reported moving many activities online to prevent the spread of the pandemic and thus shifting their strategy, which some struggled to do. Then when George Floyd was murdered there was a greater call to connect with community again, perhaps in venues not online. One PIO even reported that their position was actually created in part response to COVID-19 and in part response to the murder of George Floyd, in order to increase relations with the public. The murder of George Floyd represented a very different high-impact case that impacted PIOs slightly differently, depending on agency size. While a majority still reported being impacted by this event, the impacts were clearly different. Due to the seriousness of this event and the direct tie to law enforcement, some PIOs felt they had to shift their tone and had to respond to anti-police rhetoric online. This new role, coupled with the earlier finding that the police no longer had the barrier of news media gatekeepers, likely put stress on PIOs. The response to this stress seemed to differ with some PIOs increasing their efforts, and others becoming discouraged to the point of removing their social media presence for a time. This is reflected in the finding that many PIOs reported posting more in response to this event as well, but social media outputs showed a decrease in number of posts and then an increase in the following years. Future research should examine the differences in PIOs who chose to remove 117 themselves from social media in response to this stress and those who chose to ramp up their media efforts. Limitations The generalizability of the findings presented in this dissertation are limited based on several factors. First, during the surveying process there were some barriers to contacting PIOs. Some of the common barriers were bounced emails based on incorrect email addresses, emails or links within the emails being blocked by agency firewalls, and agency use of an online based email form instead of providing email addresses. Some PIOs suggested they would take the survey if it could be sent in a paper form but based on the skip logic and limits of resources, this was not a feasible option. These barriers prevented contact or access to the survey but are also representative of limits to email based surveys in general. The nonresponse results also indicated that while contact may have been hampered by this, the final sample did not seem to differ significantly from those who did not respond. The response rate of this survey was about 24%. While the response rate was about what was expected given the mode, timing, and length of the survey, there could have been ways to improve the response. I could have done more multi-mode contact when emails bounced or were not delivered. With more resources, this would have been a more feasible approach instead of relying primarily on email contact. There were also no resources to be used for any sort of incentive for completing the survey. Three larger negative impacts were likely the timing, the length of the survey, and the fact that many government systems block unknown emails. The response rate was also likely impacted by current events as PIOs experienced the impact of COVID-19 and the murder of George Floyd, as was presented in the results. PIOs may not have seen the survey as a valuable way to spend their time when it was stretched thin, especially due 118 to the length of the survey. While I attempted to keep the questionnaire brief, there were some complaints about the length received. I also received confirmation from initial pilot study participants that while it was a little long, the topics warranted the questions asked. Overall, it seemed that the length may have determined some, but most were receptive to the task as it was presented honestly at the start of the survey. Based on the fairly low number of emails confirmed opened, it may be that the emails were never received in the inbox of many PIOs. Government email systems may have blocked the emails completely so that the PIO never even had the opportunity to consider taking the survey. Many recent reviews of response rates for email surveys are sector specific or centered on a specific job type. There is a lack of current research examining response rates for the criminal justice sector and police in general. More research could be done on how to engage criminal justice practitioners, police, and even PIOs specifically in web-based survey research. This study is also limited by the selection of the sample, not capturing police departments with under 100 employees. This was an intentional choice in order to capture the role of the PIO specifically and not agency media management more generally. It should be noted though and future studies may wish to survey smaller agencies to better understand their interaction with media as well. It also may be interesting to compare smaller agencies who do not have a dedicated media person on staff to see what differences may be occurring in how they interact with media. Limits of the social media data were largely based on the platform used to collect it, CrowdTangle. While this platform allowed me to collect a lot of data over a large period of time, there are limitations that should be noted. CrowdTangle only captures one small piece of the Facebook ecosystem in the form of basic interaction data, post content, and post metadata. There 119 are many more nuances that simply could not be captured by numbers alone. Further, CrowdTangle only captures public content and public pages. There were some pages that were incorrectly set up by PIOs that could not be captured through the CrowdTangle platform. CrowdTangle also states that they are only looking at the most prominent public pages so some of the smaller agencies could have also been missing due to this. Finally, CrowdTangle does not retain deleted posts so if there was any particular post deleted from an agency before the collection date, this would not be reflected. Future analysis should be conducted to understand the types of interactions taking place and what these interactions were in response to. This analysis provided a starting point through an overview of social media outcomes, rather than a deep dive. This dissertation does not fully capture the audience piece of this relationship. Instead, this study focused on the PIO side of the relationship and intended to understand production of the media. This study included very surface level analysis of a small selection of social media comments and captured social media engagement with agency posts. This analysis only captured the activities of the audience who chose to interact with the posts, while there were certainly other audience members who chose to read or view the post and not interact with a reaction, comment, or share. Future research should attempt to understand the motivation of those who do choose to interact and what may motive them to. Future research should also attempt to understand the impact of the media produced by PIOs through social media as prior studies have indicated that heavy consumers of news media may have a skewed view of police agencies and misconduct (Chermak et al., 2006; Dowler & Zawilski, 2007). This information would be valuable to PIOs and is, up to this point, not well understood. Even with a further analysis of audience responses or choice not to respond, social media also is governed but a black box of 120 algorithms that can also influence who actually sees a post in their feed. Unfortunately, platform algorithms add some unpredictability that will always limit full understanding of social media interactions among agencies and citizens. This dissertation did not seek to do a full test of organizational theory and analysis and models may have been strengthened through better measurement of complete theory. There were many variables that were not found to have relationships with the social media outputs but there is cause for further investigation through more complete theory testing. This dissertation did represent a starting point in the inclusion of organizational theory in explanation of social media use by PIOs. There has been little research examining how individual or organizational elements may impact this and this dissertation serves as a first step to such research. Future research should more fully test organizational theory and impacts on social media use to confirm and expand on the findings presented here. There are likely other factors that impacted the outputs measures that were not tested. Some analysis was impacted by missing data due to the linking of multiple data sources but not in any meaningful way and this did not eliminate too many responses from analysis. Contributions This dissertation makes contributions to our understanding of the PIO role as a whole as well as diving deeper into their use of social media. As was mentioned at the beginning of this dissertation, our understanding of the PIO was largely limited to research conducted almost 20 years ago, reflecting a different media environment. While some of the findings remain the same, the introduction of social media clearly impacted the PIO, and this study contributes to the knowledge of that. 121 Another novelty of this study was the connection of survey data to social media data. While there have been a few surveys of PIOs and a few studies examining social media data, the linking of the two has been limited to this point. To keep pushing understanding forward, multiple data sources and data types should be combined in the future and this dissertation presents an example in linking survey, social media, and administrative into one dataset. Future research may also be able to capture and code more rich social media data to provide for better understanding of the environment. Combining data sources provides a more holistic view of PIOs use of social media then has been presented before. This dissertation also took place during a particularly tumultuous period which allowed for more focus on high-profile events that impacted most of those surveyed. While these events were not planned, careful consideration was taken to make sure they were added into the survey before it went live. This contributes to the knowledge of how PIOs respond to these events but also may indicate the impact that such events have on their role. Major events require shifts in the PIO role that were able to be measured by this study. Future research should always keep in mind the context that it is taking place in. This study demonstrates that PIOs actively respond to events in their roles and do so through social media too. While this study was largely exploratory and serves to highlight multiple directions for future research, there are some potential policy implications that can be noted. Police PIOs responding to this survey were surprisingly similar in demographic makeup as PIOs were about 20 years ago. This may not be surprising when one considers the bureaucratic structure of most departments, meaning many PIOs work their way up to the position in the agency, but it is surprising when one considers the dramatic change in the scope of PIO duties. As the duties have largely shifted to social media based duties, it would be important to make sure that training and 122 hiring also reflects this shift. Social media expertise may not be receiving the weight in these decisions that is proportional to the amount of the PIO job duties it takes up. Also, hiring managers may want to look at creativity and ability to update social media skills as social media require users to keep up to date in order to stay relevant. It seems like hiring more civilian employees to innovatively think about making such connections would be worthwhile. At the same time, as access to technology is not universal, there will also be a need to think creatively for delivering messaging more broadly. The results show the beginning of a communication transformation where relationships with traditional media are less important, and opportunities for manufacturing the public face of the organization directly are greater. This reflects findings that overall, local news is on the decline (Abernathy, 2018). The importance of the traditional media will continue to decline as the entire media industry is evolving. What I think this study shows is that police departments have adapted well in using social media to communicate with the public, at least these results show efforts to do so. The finding that most PIOs prefer social media for multiple tasks highlights this newer media’s importance in their role. Although it would be important to compare reactions with other types of agencies and engagement, as it seems like there is still untapped potential in using the media. Police agencies are committed to Facebook, but it only engages a small fraction of the population regularly and its importance continues to decline, especially with young people. The next police step for police organizations will be to transform these efforts and consider other platforms for the delivery of content. As communication patterns changed, especially with the opportunities for remote engagement via Zoom and similar technologies, such changes appear to be of great opportunities to interact with members of the public. Along with this, the finding that 123 PIOs are feeling more pressure to interact with and respond directly to the public may indicate a need for greater training on such interactions. These direct interactions were not always part of the PIOs job. Prior studies indicated press releases and contact with the media were the top activities PIOs partook in (Motschall & Cao, 2007). Motschall and Cao (2007) did seem to capture the beginning of this movement by finding that much of their sample was making informal contact with the public, now it seems that that has only increased. Rather than only focusing on crafting a message to push out, PIOs must become a large part of the image themselves. Surprisingly, this study found very little respondents were part of a professional organization. Joining such an organization (like the NIOA) may increase knowledge and help police PIOs stay up to date with current trends and best practices. Prior research highlights how it is difficult to maintain effective communication without such a dedication to adaptability and focus on affordances of different media (Procter et al., 2013; Dai et al., 2017). It is crucial that PIOs and agencies recognize this new demand and importance of the role. PIOs should also focus on increasing interactions with a diverse range of people. Prior literature finds that police agencies tend to be concerned about the risk of heightened visibility through social media (Dekker et al., 2020) and this study found similar sentiments reflected in the barriers and drawbacks of social media use noted by PIOs. PIOs are concerned about the increased demand for information, loss of control of a message, and responding to the public. While not everyone online will respond in a productive manner, police agencies and PIOs should attempt to increase their comfort with online interactions. Police agencies and PIOs should also attempt to reflect the diversity of their community in their agency communications. PIOs have moved from a push out information model to one that requires much more interaction and interactivity. As was suggested in previous literature, PIOs are called to better understand and 124 connect with their audience in a way that was not previously required (Hughes & Palen, 2012). This was also highlighted in the findings of this study in response to major events, which indicated that agencies need to stay flexible and reactive to their community needs, even in an online space. Future Research This dissertation worked to lay important groundwork in the police PIO space. Some future directions were presented in the finding section but overall giving PIOs opportunity to discuss their role and the intricacies about it proved to be fruitful. Some PIOs expressed their appreciation of being able to have a voice. One PIO commented that the survey “really made me reflect and think about my work” and others expressed appreciation for being included. Future research should continue to solicit PIO feedback and combine this data with other sources. In order to understand the true motivation behind social media outputs, one must understand the individuals behind the posts. This was an area that had previously been lacking and while this survey did push it forward, it is important to continue to utilize multiple sources. Future research would benefit from more complete and fine-tuned measures of social media outputs as well as organizational theory elements. While this dissertation presented both, there are likely more elements that need exploring in this under-studied area. It also seems that some of the current events may have impacted findings, but this cannot be confirmed until future research examines the elements again. Moving forward, more should be done to capture the motivations behind certain posts and the motivations behind responses. It would be beneficial to now examine the public’s motivation to respond to social media posts that are posted by PIOs to understand why some posts get more interactions and other, seemingly similar, ones do not. 125 This dissertation filled an important gap, but future research should continue to examine new technology use in the role of the PIO and explore different platforms. While this study was largely focused on the use of Facebook, there are emerging platforms that may become more important as time passes. Each platform comes with inherent capabilities, limitations, and special nuances that also must be considered and may influence PIO strategy. 126 REFERENCES Abernathy, P. M. (2020). News Deserts and Ghost Newspapers: Will Local News Survive? 1– 122. https://www.usnewsdeserts.com/wp- content/uploads/2020/06/2020_News_Deserts_and_Ghost_Newspapers.pdf Berk, R., & MacDonald, J. M. (2008). Overdispersion and poisson regression. Journal of Quantitative Criminology, 24(3), 269–284. https://doi.org/10.1007/s10940-008-9048-4 Bottoms, A., & Tankebe, J. (2013). Beyond Procedural Justice: A Dialogic Approach to Legitimacy in Criminal Justice. The Journal of Criminal Law & Criminology, 102 (1), 119- 170. https://scholarlycommons.law.northwestern.edu/jclc/vol102/iss1/4 Chermak, S. (1994).Body count news: How crime is presented in the news media. Justice Quarterly, 11(4), 561–582. https://doi.org/10.1080/07418829400092431 Chermak, S., McGarrell, E., & Gruenewald, J. (2006). Media coverage of police misconduct and attitudes toward police. Policing, 29(2), 261–281. https://doi.org/10.1108/13639510610667664 Chermak, S., & Weiss, A. (2005). Maintaining legitimacy using external communication strategies: An analysis of police-media relations. Journal of Criminal Justice, 33(5), 501– 512. https://doi.org/10.1016/j.jcrimjus.2005.06.001 Chermak, S., & Weiss, A. (2006). Community Policing in the News Media. Police Quarterly, 9(2), 135–160. https://doi.org/10.1177/1098611105281630 Cobbina, J. (2019). Hands up don’t shoot: Why the protests in Ferguson and Baltimore matter, and how they changed America. New York University Press. Colbran, M. P. (2020). Policing, social media and the new media landscape: can the police and the traditional media ever successfully bypass each other? Policing and Society, 30(3), 295– 309. https://doi.org/10.1080/10439463.2018.1532426 Crank, J. P. (2003). Institutional theory of police: a review of the state of the art. Policing: An International Journal of Police Strategies & Management, 26(2), 186–207. https://doi.org/10.1108/13639510310475723 Dai, M., He, W., Tian, X., Giraldi, A., & Gu, F. (2017). Working with communities on social media: Varieties in the use of Facebook and Twitter by local police. 41(6), 782–796. https://doi.org/doi:10.1108/OIR-01-2016-0002 Dekker, R., van den Brink, P., & Meijer, A. (2020). Social media adoption in the police: Barriers and strategies. Government Information Quarterly, 37(2), 101441. https://doi.org/10.1016/j.giq.2019.101441 127 Dillman, D. A., Smyth, J. D., & Christian, L. M. (2014). Internet, Phone, Mail, and Mixed-Mode Surveys : The Tailored Design Method. John Wiley & Sons, Incorporated. http://ebookcentral.proquest.com/lib/michstate-ebooks/detail.action?docID=1762797 DiMaggio, P., & Powell, W. (1983). The Iron Cage Revisited : Institutional Isomorphism and Collective Rationality in Organizational Fields.American Sociological Review, 48(2), 147– 160. https://doi.org/10.2307/2095101 Dowler, K., & Zawilski, V. (2007). Public perceptions of police misconduct and discrimination: Examining the impact of media consumption. Journal of Criminal Justice, 35(2), 193–203. https://doi.org/10.1016/j.jcrimjus.2007.01.006 Faber, B., Budding, T., & Gradus, R. (2020). Assessing social media use in Dutch municipalities: Political, institutional, and socio-economic determinants. Government Information Quarterly, 37(3), 101484. https://doi.org/10.1016/j.giq.2020.101484 Geiger, A. W. (2019). Online news in the U.S.: Key facts | Pew Research Center. https://www.pewresearch.org/fact-tank/2019/09/11/key-findings-about-the-online-news- landscape-in-america/ British Broadcasting Corporation. (2020, June 4). George Floyd death: US protests timeline. BBC News. https://www.bbc.com/news/world-us-canada-52921418 Goldsmith, A. J. (2010). Policing’s new visibility. British Journal of Criminology, 50(5), 914– 934. https://doi.org/10.1093/bjc/azq033 Gottfried, B. Y. J., Walker, M., & Mitchell, A. (2020). Americans see skepticism of news media as healthy, Say public trust in the institution can improve. https://www.pewresearch.org/journalism/2020/08/31/americans-see-skepticism-of-news- media-as-healthy-say-public-trust-in-the-institution-can-improve/ Graham, M. W. (2014). Government communication in the digital age: Social media’s effect on local government public relations. Public Relations Inquiry, 3(3), 361–376. https://doi.org/10.1177/2046147X14545371 Graziano, L. M. (2018). News media and perceptions of police: a state-of-the-art-review. Policing, 42(2), 209–225. https://doi.org/10.1108/PIJPSM-11-2017-0134 Graziano, L. M., & Gauthier, J. F. (2018). Media consumption and perceptions of police legitimacy. Policing, 41(5), 593–607. https://doi.org/10.1108/PIJPSM-12-2016-0177 Hitlin, P., & Vogt, N. (2014, August 20). Cable, Twitter picked up Ferguson story at a similar clip | Pew Research Center. https://www.pewresearch.org/fact-tank/2014/08/20/cable- twitter-picked-up-ferguson-story-at-a-similar-clip/ 128 Hu, X., Rodgers, K., & Lovrich, N. P. (2018). “We Are More Than Crime Fighters”: Social Media Images of Police Departments. Police Quarterly, 21(4), 544-572. https://doi.org/10.1177/1098611118783991 Hu, Xiaochen, & Lovrich, N. P. (2020). Electronic community-oriented policing: Theories, contemporary efforts, and furture directions. Lexington Books. Hughes, A. L., & Palen, L. (2012). The Evolving Role of the Public Information Officer: An Examination of Social Media in Emergency Management. Journal of Homeland Security and Emergency Management, 9(1). https://doi.org/10.1515/1547-7355.1976 Intravia, J., Wolff, K. T., & Piquero, A. R. (2018). Investigating the Effects of Media Consumption on Attitudes Toward Police Legitimacy. Deviant Behavior, 39(8), 963–980. https://doi.org/10.1080/01639625.2017.1343038 Jefferis, E. S., Kaminski, R. J., Holmes, S., & Hanley, D. E. (1997). The effect of a videotaped arrest on public perceptions of police use of force. Journal of Criminal Justice, 25(5), 381– 395. https://doi.org/10.1016/S0047-2352(97)00022-6 Justice, U. S. D. of. (2016). Law Enforcement Agency Roster (LEAR). Interuniversity Consortium for Political and Social Research. https://doi.org/https://doi.org/10.3886/ICPSR36697.v1 Kääriäinen, J., Isotalus, P., & Thomassen, G. (2016). Does public criticism Erode trust in the police? The case of Jari Aarnio in the Finnish news media and its effects on the public’s attitudes towards the police. Journal of Scandinavian Studies in Criminology and Crime Prevention, 17(1), 70–85. https://doi.org/10.1080/14043858.2016.1144315 Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons, 53(1), 59–68. https://doi.org/http://dx.doi.org/10.1016/j.bushor.2009.09.003 Kim, K., Oglesby-Neal, A., & Mohr, E. (2017). 2016 Law Enforcement Use of Social Media Survey. Urban Institute. https://www.urban.org/sites/default/files/publication/88661/2016- law-enforcement-use-of-social-media-survey_5.pdf Lieberman, J. D., Koetzle, D., & Sakiyama, M. (2013). Police Departments’ Use of Facebook:Patterns and Policy Issues. Police Quarterly, 16(4), 438–462. https://doi.org/10.1177/1098611113495049 Mawby, R. C. (2010). Police corporate communications, crime reporting and the shaping of policing news. Policing and Society, 20(1), 124–139. https://doi.org/10.1080/10439461003611526 Medaglia, R., & Zheng, L. (2017). Mapping government social media research and moving it forward: A framework and a research agenda. Government Information Quarterly, 34(3), 129 496–510. https://doi.org/10.1016/j.giq.2017.06.001 Meijer, A., & Thaens, M. (2013). Social media strategies: Understanding the differences between North American police departments. Government Information Quarterly, 30(4), 343–350. https://doi.org/http://dx.doi.org/10.1016/j.giq.2013.05.023 Meyer, J. W., & Rowan, B. (1977). Institutionalized Organizations : Formal Structure as Myth and Ceremony. The American Journal of Sociology, 83(2), 340–363. https://doi.org/10.1086/226550 Minneapolis Police’s Public Information Officer To Be Under City’s Supervision. (2020, September 23). Retrieved November 20, 2020, from https://minnesota.cbslocal.com/2020/09/23/minneapolis-polices-public-information-officer- to-be-under-citys-supervision/ Motschall, M., & Cao, L. (2007). An analysis of the public relations role of the police public information officer. 5(15), 1–25. https://doi.org/10.1177/109861102129198084 Nix, J., & Wolfe, S. E. (2017). The Impact of Negative Publicity on Police Self-legitimacy. Justice Quarterly, 34(1), 84–108. https://doi.org/10.1080/07418825.2015.1102954 Nix, J., & Wolfe, S. E. (2018). Management-level officers’ experiences with the Ferguson effect. Policing, 41(2), 262–275. https://doi.org/10.1108/PIJPSM-11-2016-0164 Nix, J., Wolfe, S. E., & Campbell, B. A. (2018). Command-level Police Officers’ Perceptions of the “War on Cops” and De-policing. Justice Quarterly, 35(1), 33–54. https://doi.org/10.1080/07418825.2017.1338743 Rabe, M., & Jensen, E. (2023). Exploring the racial and ethnic diversity of various age groups. https://www.census.gov/newsroom/blogs/random-samplings/2023/09/exploring- diversity.html Rainey, H. (2003). Understanding and managing public organizations (3rd ed.). John Wiley & Sons, Incorporated. Roche, S. P., Pickett, J. T., & Gertz, M. (2016). The Scary World of Online News? Internet News Exposure and Public Attitudes Toward Crime and Justice. Journal of Quantitative Criminology, 32(2), 215–236. https://doi.org/10.1007/s10940-015-9261-x Rodgers, K. (2016). A Content Analysis of Social Media Policies of Police Departments and Sheriffs' Offices in the State of Idaho: Congruency to the Model Policy [Master's Thesis, Boise State University]. Boise State University ScholarWorks. https://scholarworks.boisestate.edu/cgi/viewcontent.cgi?article=2147&context=td Simpson, R., & Wetherell, K. (2020). Police spokespersons: A question of organizational representation. International Journal of Police Science & Management, 22(3), 253–262. https://doi.org/10.1177/1461355720918882 130 Skolnick, J. H., Mccoy, C., American, S., Foundation, B., Summer, N., Skolnick, J. H., & Mccoy, C. (1984). American Bar Foundation Police Accountability and the Media Linked references are available on JSTOR for this article : Police Accountability and the Media. 9(3), 521–557. Stocking, G., van Kessel, P., Barthel, M., Matsa, K. E., & Khuzam, M. (2020). Many Americans Get News on YouTube, Where News Organizations and Independent Producers Thrive Side by Side | Pew Research Center. https://www.journalism.org/2020/09/28/many-americans- get-news-on-youtube-where-news-organizations-and-independent-producers-thrive-side-by- side/ Surette, R., & Richard, A. (1995). Public information officers: A descriptive study of crime news gatekeepers. Journal of Criminal Justice, 23(4), 325–336. https://doi.org/10.1016/0047- 2352(95)00023-J ABC News. (2020, September, 22). Timeline: How coronavirus got started. Retrieved December 6, 2020, from https://abcnews.go.com/Health/timeline-coronavirus- started/story?id=69435165 Walker, S. (2015). Sense and nonsense about crime, drugs, and communities (8th ed.). Cengage Learning. 131 APPENDIX Table A.1 Updating understanding of PIOs survey items Research Question Who are the personnel responsible for marketing police? Survey Items How many people in your organization are considered public information officer staff? (How many staff positions are in the PIO office?) How many civilians? How many sworn? Does the PIO staff manage media inquiries from outside organizations? Does the PIO staff manage department social media accounts? In a typical month, how often does your office hold meetings? Demographics: What is your employment status within the department? (Please circle the most appropriate response/s) Civilian a) Sworn b) c) Other (specify: ) What is your rank in the department? a) b) c) d) e) f) g) h) i) Officer Corporal Detective Sergeant Lieutenant Captain Commander Not applicable or unknown Other (specify: ) How long (in years) have you served in your current position? ________ What gender do you identify as? a) b) c) d) e) Female Male Non-binary or Transgender Other Prefer not to say 132 Table A.1 (Cont’d) What is your race? (Select all that apply) a) b) c) d) e) f) g) Asian American Indian or Alaskan Native Black or African American White Unknown Other Prefer not to say Are you of Hispanic or Latino/a/x Origin? a) b) c) d) e) f) No, not of Hispanic, Latino/a/x origin Yes, Mexican, Mexican American, Chicano/a/x Yes, Puerto Rican Yes, Cuban Yes, other Hispanic, Latino/a/x origin Prefer not to say What is your age (in years)? What type of training do personnel responsible receive Educational Background Occupational Background About how many hours of formal training have you received in police- news media relations in the past 12 months? Who in department receive media training? What types of formal training have you received in media communication? (Check all that apply) A) Educational background (i.e., degree in journalism, public relations, etc.) B) Prior occupational background (i.e., working in media/public relations) C) In-service Training D) Other (Please specify) _____________ E) None Would you say the amount of formal training, provided by the department, you have received in police-news media relations is: (please circle one) Poor Excellent Good Fair N/A Would you say the amount of formal training, outside of the department, you have received in police-news media relations is: (please circle one) Poor Fair Good Excellent N/A 133 Table A.1 (Cont’d) What are the current communication methods being used by police PIOs in the U.S.? What strategies are being used by PIOs to market police in media? Does your agency have a written media policy? Year adopted Year(s) updated In a typical month, how many press releases does your office issue? In a typical month, how many news conferences does your office hold? In a typical month how many community meetings does somebody in your office attend? In a typical month how often do you communicate with other PIOs? In a typical day, how many calls/emails/inquiries do you receive from local new organizations? How frequently do you meet with local news media leadership (editors/managers)? a) Daily b) Weekly c) Monthly d) Yearly e) Do not meet For each of the following issues, what is the preferred media outlet (News paper, Radio, TV, Social Media, Other (Specify): Request information from public Build Relationships with the public Report Crime Incident Information Report Crime Arrest Information Discuss New Police Programs Discuss New Police Policies. Report Department Awards and Promotions Respond to Critical Assessments. What are the relationships like between media and PIOs? What are the PIOs perceptions of portrayal of police by different media? In a typical day, how many personnel from the following might you interact with: a. Newspapers _____ (number) b. Television stations _____ (number) c. Radio stations ______ (number) d. Other media ______ How would you describe the quality of your department’s working relationship with news reporters? Poor Fair Good Excellent N/A 134 Table A.1 (Cont’d) In general, how would you describe the quality of your department’s image currently within the local news media? Poor Fair Excellent Good N/A Table A.2 PIOs and Social media survey items Research Question Concepts Survey Items How do PIOs utilize social media differently than traditional media? methods used to market police departments on social and traditional media Do you and your staff use social media for public information duties? a. Yes b. No c. Not Sure If yes, what type? Facebook Twitter Instagram i. ii. iii. iv. Youtube v. Other (list) ___________ What social media platform does your department use most often? (Select One) i. Facebook Twitter Instagram ii. iii. iv. Youtube v. Other (list) ___________ Which do you consider to be most important? ________________ Who has routine authority to communicate using department Social Media (Check all that apply)?  PIO staff  Chief  Command Staff  Officers  Other Civilian Staff Who develops messages for social media (Check all that apply)?  PIO staff  Chief 135 Table A.2 (Cont’d) What strategies are being used to communicate with the public, do they vary based on format (social media vs. traditional media)? methods used to market police departments on social and traditional media  Command Staff  Officers  Other Civilian Staff Does your agency have policies regulating the information that you put out through traditional media (television, newspapers, radio)? a. Yes b. No c. Not Sure d. Policy is in development Does your agency have policies regulating your use of the department’s social media pages? a. Yes b. No c. Not Sure Policy is in development How many social media posts do you or your staff make in a typical month? How many press releases do you or your staff produce in a typical month? Do you complete any other information producing activities in a typical month? ___________________________ How would you describe your typical social media post? In a typical month, how often do you or your staff do each of the following? Communicated through social media with community groups Contacted a social media company for assistance In your opinion what is the best way to publicize each of the following: Community policing efforts Successful criminal investigation (leads to arrest) Request for information about a case Departmental promotions/New Hires Department policies New initiatives 136 Table A.2 (Cont’d) What changes has social media brought to police communication, specifically, what are the benefits and drawbacks? perceived benefits and drawbacks of using social media, d Options include a. Television b. Newspapers c. Radio d. Social media Other (please specify) _______________ What is the biggest benefit social media has brought to your department? a. Connection with community members b. Quick way to notify public of safety concerns c. Public relations and reputation management d. Recruitment and applicant vetting e. Communicating with other government agencies f. Soliciting tips on crime from public g. Intelligence gathering for criminal investigations h. Other (please specify) _______________ What is the biggest drawback social media has brought to your department? a. Time spent developing content for social media b. Increased visibility of department c. Increased demand for quick information from public d. Increased criticism from public through social media e. Loss of control of the meaning of the message f. Other (Please Specify) _____________ What are the barriers you have faced in using social media for your department (Select all that apply)? a. Gaining agency leader’s support b. Adapting to new trends c. Measuring the impact of social media d. Balancing timeliness with accuracy e. Training personnel to use social media effectively f. Building and managing IT infrastructure needed for social media g. Lack of department resources devoted to social media 137 Table A.2 (Cont’d) training and knowledge of various social media platforms What department resources (training, funding, time) are being put towards social media and how do PIOs collaborate with other units in the organization to share or collect information via social media? h. Using social media in a way that does not create a liability i. Designing policy for department social media use j. Responding to information received from the public on social media k. Other (please specify) _______________ What types of formal training have you received in social media specifically? (Check all that apply) a. Educational background (i.e., degree in journalism, public relations, etc.) b. Prior occupational background (i.e. working in media/public relations with social media) c. In-service Training d. Other (Please specify) _____________ e. None About how many hours of formal training have you received on social media use in the past 12 months? Would you say the amount of formal training you have received in use of social media for your department in the past is: (please circle one) Poor Fair Good Excellent N/A Has your department put any of the following resources towards the use of social media for public information activities? a. Funding sure ▢ yes ▢no ▢ not b. More staff ▢ yes ▢ no ▢ not sure c. Dedicated time ▢ yes ▢ no ▢ not sure Other resources (please specify)_____________ Do you work with other units within your organization to release information on social media? a. b. c. Not Sure Yes No IF yes, which ones? ____________ 138 Table A.2 (Cont’d) Do organizational variables (e.g., size of the police organization, budget, extant policy, leadership; technological support/capacity) impact the PIOs perceptions of the value of social media? Does it impact their expertise and skills in using social media? Do these variables impact the activity and engagement of social media use by an organization? Do individual level PIO variables (understanding of the platform, training and experience with social media, motivation) impact activity and engagement of social media use? Do you utilize the Facebook Insights on your agency page? a. Yes b. No c. Not Sure training and knowledge of various social media platforms What is your department’s approximate yearly budget allocated specifically to PIO duties? Does upper-level management in your department support the department’s use of social media? a. Yes b. No c. Not Sure Does your agency have technological support to assist with your use of social media for the department? c. Yes d. No e. Not Sure the personnel responsible for this work and training and knowledge of various social media platforms How important do you feel social media is to your position as a PIO? a. Very Important b. Important c. Moderately Important d. Slightly Important e. Not Important How much you enjoy using social media as your department PIO? a. Very much enjoy b. Enjoy c. Somewhat Enjoy d. Slightly Enjoy e. Not at all Enjoy 139 Table A.2 (Cont’d) Are organizational or individual level variables better predictors activity and engagement of social media use? How competent do you feel using social media for your job as a PIO? a. Very competent b. Competent c. Somewhat competent d. Slightly competent e. Not at all competent How long (in years) have you used social media in your job as a PIO? How long (in years) have you personally used social media? How often do you use social media in your personal life? a. more than once a day b. daily c. weekly d. monthly e. less often What social media platforms do you use in your personal life? a. Facebook b. Twitter c. Instagram d. Youtube e. Other (list) ___________ What social media platform do you use most often in your personal life? (Select One) a. Facebook b. Twitter c. Instagram d. Youtube e. Other (list) ___________ How long has your organization maintained a social media page? see social media data Size of agency Population served Number of sworn and civilian 140 Table A.3 Events of 2020 survey items Research Question How has COVID- 19 impacted PIO strategy and use of social media, if at all? Concepts Survey Items e) public relation responses to the COVID-19 pandemic Was your job as a PIO impacted by COVID-19? a. Yes, a lot b. Yes, a little c. No, not at all d. Not sure If so how? Did your relationship with news media change as a result of COVID-19? a. Yes, it became worse b. Yes, it became better c. No, it did not change d. Not sure Did your use of social media for your department change as a result of COVID-19? a. Yes, we used it more b. Yes, we used it less c. No, our use did not change d. Not sure If yes, we used it more, how did you use social media differently? (check all that apply) a) Department posted more frequently b) Responded more frequently to public’s questions c) Held more virtual events d) Posted more public health information e) Other (Please specify)___________ 141 Table A.3 (Cont’d) How has the death of George Floyd impacted PIO strategy and use of social media, if at all? f) public relation responses following the death of George Floyd. How did each of these events impact social media engagement overtime, if at all? Was your job as a PIO impacted by the killing of George Floyd and/or subsequent protests? a. Yes, a lot b. Yes, a little c. No, not at all d. Not sure If so, how? Did your relationship with news media change after the killing of George Floyd and/or subsequent protests? a. Yes, it became worse b. Yes, it became better c. No, it did not change d. Not sure Did your use of social media change after the killing of George Floyd and/or subsequent protests? a. Yes, we used it more b. Yes, we used it less c. No, our use did not change d. Not sure If yes, we used it more, how did you use social media differently? (check all that apply) a) Department posted more frequently b) Responded more frequently to public’s questions c) Held more virtual events d) Other (Please specify) ___________ N/A – see social media data 142