Introduction

The city of Flint is a typical example of the de-industrialization and economic decline experienced by several U.S. cities after decades of prosperity in the last part of the twentieth century ^{(Bluestone and Harrison, 1982}; ^{High and Lewis, 2007)}. Flint was an icon of the U.S. automobile industry as the place where General Motors (GM) Corporation was originally founded. The city expanded from the mid-1940s until the late 1970s, as gm generated thousands of auto manufacturing jobs for unskilled workers. However, from the 1980s through most of the 1990s, Flint began its steep decline due to closures and the relocation of thousands of General Motors manufacturing jobs ^{(Highsmith, 2014)}. As a result, Flint has lost population and experienced high unemployment rates and low income levels, compared not only to the rest of Michigan but to the rest of the country.

Flint’s socio-demographic indicators show the social and economic impacts. First of all, total population declined from 1960 to 2010 by almost half (48 percent). However, the percentage of black population increased from 17.5 percent in 1960 to more than half (56.6 percent) in 2010 ^{(U.S. Census Bureau, 2015)}. Secondly, Flint unemployment rates increased in just ten years by 7 percent, from 16.1 percent in 1990 to 23.2 percent in 2010, which was higher than the state as a whole and the metropolitan area (MSA) ^{(Doige et al., 2015: 8)}.

And thirdly, the average annual household income from 2010 to 2014 in Flint was US$24,679, which is much lower than the state level of US$49,087, and less than half that of the national level of US$53,482 ^{(U.S. Census Bureau, 2015)}. Closely linked to this figure is the percentage of the population living below the poverty level, which has increased from 16.9 percent in 1980 to 31.1 percent in 2010; this is more than double the rate for Michigan (15.5 percent) and the country as a whole (15.1).

As result of these conditions, tax revenues declined while local services continued to cost the city money, causing significant fiscal stress to the Flint local government ^{(Scorsone and Bateson, 2011: 1}; ^{Doige et al., 2015: 13)}. The city was so financially weak that Governor Rick Snyder appointed an emergency manager (EM) on November 8, 2011,¹ with enough power to determine the city’s financial functioning. This displaced democratically-elected public officials by appointing a non-elected official to manage the city.² The functions and operations of the EM as conceived generate tensions in the traditional democratic principles of checks and balances ^{(Kasdan, 2014)}.

Flint’s financial situation is part of the deepest roots of the Flint water crisis. To save money, the em made key decisions that seem to have contributed to the crisis, “from the use of the Flint River to delays in reconnecting to the Detroit Water System (DWS) once water quality problems were encountered” ^{(FWATF, 2016: 1)}.

Our research shows that negative conditions coincided at a time when residents were unaware of the quality of their water supply. Government institutions and the news media both failed to respond and inform residents in an accountable and transparent way of the harmful water conditions, although the former were responsible for providing safe public services to all citizens and the latter, for watching over the public function and informing the public of any irregularities.

The Underlying Conditions of the Flint Water Crisis

The Flint water crisis evolved as a trickle, not as a single crisis event. Residents reported problems soon after the city switched to its own water source on April 25, 2014. They complained about the odor, taste, and color of the water coming directly into their homes; and they began to report health complaints, such as skin rashes ^{(Hanna-Attisha et al., 2016)}. Throughout this time city and state officials assured residents that the water was safe to drink ^{(Gordy, 2016}; ^{Veolia North America, 2015)}.

The problems with Flint’s new water source did not begin with high levels of lead, but rather the city’s efforts to disinfect it. Contaminants entered the water supply in August 2014 and boil-water advisories followed. City officials blamed the first on an “abnormal test,” but after two additional advisories in September, they cited probable weaknesses in the infrastructure that were allowing bacteria to enter the system ^{(Adams, 2014)}. They added more chlorine, a highly corrosive substance, to remove the contaminants ^{(Al-Jasser, 2007}; ^{Fonger, 2014)}. In the early stages of the crisis, the Flint Journal (MLive) described water treatment efforts as “a crash course in learning how to treat river water” ^{(Fonger, 2014)}.

During this time, water treatment staff members were already struggling with disinfection. In May and August, sampling revealed excessive levels of a chemical byproduct called trihalomethane, or TTHM, which is created when chlorine reacts to organic matter ^{(FWATF, 2016}; ^{Sivaganesan and Clark, 1998}; ^{TTHM Notice, 2015)}. Levels were so high that they violated the federal standard for the Safe Water Drinking Act. In January 2016, the city was forced to notify residents that the sample, which is averaged over six months, showed a violation. The notice stated the water was not a health risk for “most people,” and that they had already implemented measures to improve water quality after the May and August testing ^{(TTHM Notice, 2015)}.

The notices served as a catalyst for activism throughout Flint, which attracted the attention of a well-known environmentalist ^{(Fonger, 2015a)}. She offered to perform a free evaluation of the city’s water supply, but city officials declined and contracted Veolia North America ^{(FWATF, 2016}; ^{Veolia North America, 2015)}. Veolia’s report concurred with the city: the water was safe coming out of the plant, and any discoloration was the result of cast-iron erosion from pipes. The consultant’s action plan recommended improved treatment methods, staff training, and a new carbon filtration system ^{(Fonger, 2015b}; ^{Veolia North America, 2015)}. The report failed to mention possible erosion in homes and service lines ^{(FWATF, 2016}; ^{Veolia North America, 2015)}.

Complaints about the color, odor, and taste continued until a group of outsiders brought national attention to what government officials and the media failed to address: residents were ingesting contaminated water as a result of corrosion ^{(Roy, 2015a}; ^2015b). First, ^{Roy (2015a)} released a study by Virginia Tech researchers on August 31, 2015, which indicated that almost half the samples they collected from 48 homes in Flint contained dangerous levels of lead. E-mails later revealed that some governmental officials, including representatives from the Department of Health and Human Services and the Michigan Department of Environmental Quality, knew of the increased lead levels, but failed to act, marginalized the situation, and attempted to conceal evidence ^{(FWATF, 2016: 16}; ^{Roy, 2015}; ^{Roy and Edwards, 2016}; ^{Michigan Government, 2016)}.

The second major catalyst brought worldwide attention to the city of Flint and its water problems. A report by Michigan State University researchers ^{Mona Hanna-Attisha, Jenny LaChance, Richard Casey Sadler, and Allison Champney Schnepp (2016)} released in late September 2015 found that the incidence of lead in children’s blood in the city of Flint had increased from 2.4 percent to 4.9 percent after the water source changed. The city would also later report an increase in deaths from Legionnaires’ disease after the switch ^{(Burke, 2016}; ^{FWATF, 2016}; ^{Hanna-Attisha and Kuehn, 2016)}. The people of Flint were poisoned by the city’s water supply, carrying unknown long-term effects such as behavioral and learning problems and reductions in I.Q. ^{(Hanna-Attisha et al., 2016}; ^{U.S. EPA, 2017)}.

On December 14, 2015, the city declared a state of emergency. The state of Michigan followed suit three weeks later ^{(FWATF, 2016)}. The water’s damage potential, which General Motors recognized to be so corrosive to auto parts that it had disconnected from the city’s water supply more than a year previously, was blamed for eating away at the service lines and household plumbing of the city’s 100 000 water customers ^{(Fonger, 2014d}; ^{FWATF, 2016)}, potentially exposing them to metals including lead and copper ^{(U.S. EPA, 2016}; ^{Kammer, 2015}; ^{City of Flint, 2015)}.

The water crisis continues to attract attention. Criticisms are directed at the federal, state, and local governments, the media, the consulting firm, and even the local hospitals for not doing enough to protect the citizens of Flint ^{(Halcom, 2016}; ^{Hearing Wrap-Up, 2016}; ^{Robbins, 2016)}. Both state and federal officials called on Michigan’s governor to resign. Government officials face criminal and misdemeanor charges of tampering with test results and making false statements about water testing ^{(Clark, 2016}; ^{Michigan Government, 2016}; ^{Bitley, n. d.)}.

The Flint water crisis is a unique case in which the public mobilized to draw attention to an ongoing public health threat that continued for more than a year with little resolution. Questions remain as to what happened throughout the period leading up to the discovery of lead in the water and what actions and players are responsible for the state of affairs remaining hidden for so long.

Literature Review

In the last few years, scholars and practitioners have increasingly focused on -or have returned their attention to- open government ^{(Bertot, Jaeger, and Grimes, 2010}; ^{Luna-Reyes, Bertot, and Mellouli, 2014)}. However, its study is not new and the concept has been used in different contexts ^{(Parks, 1957}; ^{Nam, 2012)}. Bernd Wirtz and Steven Birkmeyer formulate a comprehensive concept that includes the essence of the peer-review literature from 2008 to 2013 that they examined. For them,

Open government is a multilateral, political, and social process, which includes in particular transparent, collaborative, and participatory action by government and administration. To meet these conditions, citizens and social groups should be integrated into political processes with the support of modern information and communication technologies, which together should improve the effectiveness and efficiency of governmental and administrative action. ^{(Bernd Wirtz and Steven Birkmeyer, 2015: 382-384)}

^{Wirtz and Birmeyer (2015)} underline the importance of transparency as one of the most essential bases for open government. Transparency has been a part of U.S. American ideology since the beginning of its democracy ^{(Harder and Jordan, 2013)}. In this sense, the right to access government information is seen as a way to facilitate democratic participation, prevent corruption, and establish trust in government, providing information to citizens, particularly journalists. The U.S. Congress passed the Freedom of Information Act (FOIA) in 1966 to provide a more specific legal support for transparency.

Transparency shares a special relationship with accountability because, simply, “a democracy requires accountability and accountability requires transparency” ^{(White House, 2016)}. Public accountability is a cornerstone of good governance and, again, a basic principle of democracy (for example, see ^{Bovens, Schillemans and Hart, 2008}; ^{Mulgan, 2000}; ^{Aucoin and Heintzman, 2000}; ^{Przeworski, Stokes, and Manin, 1999)}.

Accountability must be seen as a process. In fact, many scholars have highlighted three analytical steps or phases (for example, ^{Schillemans, Van Twist, and Vanhommerig, 2013}; ^{Mulgan, 2003)}. The first is the information phase, which presupposes that an accountable entity or public official provides retrospective information about her/his/its choices, conduct, or performance. The second is the debate phase, when the accountability forum evaluates the information received and starts a debate with the accountable entity or public official. And third is the sanctions phase, where accountability permits some kind of judgment on the activity of the political official or entity and leads to punishment or reward for his/her/its actions. The most significant kind of accountability is produced by agencies that can reveal existing data, investigate, and produce information about institutional behavior ^{(Fox, 2007: 668-669)}. This raises the question of what happens when public officials fail to fulfill their mandate.

As the fourth estate, the press has seen itself as a form of accountability, because the “ideal goal of traditional journalism has been to make power accountable: to keep ordinary citizens apprised of what government is doing, and how it affects them both individually and with respect to the groups and values that they care about” ^{(Entman, 2005: 48)}. From the moment that the media pay attention to governmental failures or political abuses and promotes social debate about them, the press is perceived as a way to strengthen accountability (see, for example, ^{Ettema, 2007}; ^{Chavez, 2006)}. The news media provide a way for citizens to see how public officials respond to their problems ^{(Maia, 2009)}. This, in turn, raises two questions: what happens when the media do not report deeper public affairs and simply reproduce official messages? And, where is the watchdog function that is supposed to be vigilant at all times?

One important function of the press is to monitor government actions and the impacts they have on the community. This is called the media watchdog function. It is critical since the press is touted as serving the public and the communities they cover. The function of the media watchdog is well established within normative theory. ^{Laswell (1948)} considered the watchdog function to be one of surveillance, monitoring the environment, collecting evidence, and reporting relevant facts to the public they inform. Scholars have expanded on this premise to suggest the media remain not exclusively objective, but operate with a critical eye, evaluating sources and making value judgments in order to give voice to citizens and act as an advocate if needed ^{(Ettema and Glasser, 1998}; ^{Glasser, 1989)}. This function holds journalists accountable to expose corruption in the government or in the private sector and to hold leaders accountable with regard to their responsibilities. ^{(Ettema and Glasser, 1998}; ^{Laswell, 1948}; ^{Nielsen, 2015)}. The media receive criticism when they fail to serve this purpose, and are sometimes referred to as being a subservient “lapdog” to the government or special interests, or perhaps playing an insignificant role in society ^{(Christians, McQuail, and Glasser et al., 2009}; ^{Dennis, Gilmor, and Glasser, 1989}; ^{Johnson, 2014}; ^{Donohue, Tichenor, and Olien, 1995)}.

The public rates watchdog reports focusing on subjects such as corruption and accountability among the top five most important investigative areas. However, recent economic and educational factors are blamed for slashing investigative budgets in both professional and academic settings ^{(Clark, 2013}; ^{Houston, 2010}; ^{Wasserman, 2008)}. Newsrooms cut staff by approximately 25 percent between 2000 and 2009 ^{(Starkman, 2014)}. Therefore, editors must efficiently maximize content with minimal staff often resulting in less priority being placed on “working stories” and those requiring follow-up ^{(Jones, 2013)}. It makes sense that the media are most successful at uncovering lack of transparency and corruption in times of immediate crisis, or big stories, when more resources are dedicated to in-depth, continuing coverage of an event, rather than a slow-to-develop crisis, as was the case in Flint.

Communities expect local journalists to hold government officials and sources accountable, to be solution-driven and connected to the communities they serve. Journalists often fall short, opting for the quick, “official” government, elite, or business source ^{(O’Neill and O’Connor, 2008}; ^{Starkman, 2014)}. They often fail to develop a story with multiple viewpoints and angles from across the socio-economic spectrum. As a result, citizens and activists are much less frequently cited then government officials or businesses ^{(O’Neill and O’Connor, 2008}; ^{Neilsen, 2015)}. In terms of the Flint water crisis, which in its early stages was best characterized by citizen activism without government response, previous literature regarding news frames involving citizen movements may offer further insight. Historically, U.S. newspaper coverage favors the status quo and provides less voice to the disenfranchised; this holds true even in non-violent situations (Boyle, McLeod, and Armstrong, 2012; McLeod and Armstrong, 2012; Armstrong, Boyle, and McLeod, 2012). Activism and citizen-focused stories are less frequent than those with a governmental focus, especially in local news ^{(Nielsen, 2015}; ^{O’Neill and O’Connor, 2008)}.

Recent case studies find that the media watchdog can receive both praise and criticism for its crusade against a non-transparent government. In cases in which official sources lag behind or fail to respond adequately, the media turn to independently exploring answers. This was the case of Hurricane Katrina, when the government was slow to react (Garnett and Kouzmin, 2007), and the Elk River chemical spill in Charleston, West Virginia, when confusing information and a lack of accountability by government officials led the media to use technology to pursue their own answers ^{(Bray, 2015)}. Both, in turn, resulted in errors in content.

The case of the water crisis in Flint, Michigan, may demonstrate failures in the both the media watchdog function and governmental institutions over the period leading up to the exposure of lead leaching into the homes of the city’s water. From the moment when the city switched its water source to the Flint River until researchers brought the lead crisis to light, citizens complained of health effects, tested positive for poisoning from metals, and brought their contaminated water to city officials ^{(Fonger, 2015c}; ^{Gordy, 2016}; ^{Hanna-Attisha et. al, 2016)}. At the same time, e-mail evidence and other testimony reveals officials at various levels of the government manipulated and attempted to conceal safety testing results ^{Hearing Wrap-Up, 2016}; ^{Roy, 2015a}; ^{Roy and Edwards, 2016)}.

Data Collection and Methodology

The vast majority of the literature on accountability emanates from political science and public administration and pays scant attention to the role of the news media. Obviously, this point of view is completely different in media studies, which presents the watchdog function as essential; however, studies on accountability from this field of knowledge are by no means abundant (Djerf-Perre, Ekström and Johansson, 2013). That is why we focused on the nature of the how institutions informed and delivered actions to the Flint community. To examine these issues, we used qualitative interdisciplinary approaches that provided a framework linking accountability and transparency across institutions. The research design sought to answer the following questions:

RQ1: Was there a major institutional failure during the Flint water crisis?

RQ2: What was the thematic coverage of the news media during the Flint water crisis?

RQ3: What was the role of the media in the initial stages of the Flint water crisis?

The Flint water crisis is a fluid process that is expected to linger for years to come. This study focuses on the early stages of the crisis until President Obama declared the State of Emergency; it uses a proxy of the newspapers that represent the coverage at the national, state and, local levels. To identify the relevant dates during the Flint water crisis, the researchers used the Flint Water Advisory Task Force Final Report ^{(FWATF, 2016)} to pinpoint how the media reported on important events.

To analyze the function of the national and state media we collected New York Times articles from the Factiva database and used ProQuest to search the Detroit Free Press. In terms of The Flint Journal (MLive), we selected the online version because the print version is not a daily newspaper. For this source, researchers ran a day-by-day Google News search utilizing the search terms, “Flint” and “Water.” The search was rerun utilizing the names of The Flint Journal (MLive) staff writers to locate additional articles. Also, the media stories and timelines of events were cross-referenced to identify the level of government response, actions, and approaches during the crisis. A discretionary sample of this kind of cross-referencing is displayed in Appendix A.

The initial period of examination began September 2, 2015, the date of the first detailed news coverage about a Virginia Tech report of lead leaching into the city’s water supply, and ended on January 16, 2016, when President Barack Obama declared a State of Emergency in Flint ^{(White House, 2016)}. The New York Times did not reveal significant samples for analysis, so the sample was expanded to include articles from January 1, 2015 through January 17, 2016.

The parameters required source-specific content pertaining only to the issues relating to and resulting from Flint’s water contamination. Wired and press bureaus such as the Associated Press articles and editorials written by an external source were excluded (Table 1). The total sample resulted in 253 articles (n =253).

The qualitative content analysis consisted of coding key data into themes through an inductive process, which is more conducive to categorization when unpredictable themes might be expected to emerge (Lindlof and Taylor, 2011). We analyzed articles through a four-step process to develop key themes in the articles published by the national, state, and local media. First, we read the articles for overall content and summarized them. Our analysis then considered tone of headlines, quotes, and the overall article. The third step involved generating descriptive codes utilized within each article. ^{Graneheim and Lundman (2004)} suggest this type of extensive categorization makes an analysis of underlying meanings within texts more reliable. The main themes for each source were ranked in descending order (Table 2).

Definitions of themes:

Findings

“The Flint water crisis is a story of government failure, intransigence, unpreparedness, delay, inaction, and environmental injustice.” ^{(FWATF, 2016: 1)}

Discussion and Limitations

Further, frames within national, state, and local reporting remained consistent with previous scholarly findings regarding accountability and the media’s watchdog role. ^{Bennett (2007)} finds that national news often provides broader perspectives of contentious political issues. In terms of how political and other official sources drove local media content, the findings support research indicating that local coverage may favor political and organized groups with established influence within the community ^{(Donohue, Tichenor, and Olien, 1995}; ^{Davis, 2009}; Tichenor et al., 1980). ^{Bennett (1990)} and ^{Hallin (1994)} find that reliance on political sources and their public relations officials makes coverage less critical. In this sense, ^{Balkin (1999)} states that mass media can help simulate political transparency, not by hiding facts, but by altering the context in which citizens view them. This is close to what the local newspaper did in Flint water crisis.

We must point out the limitations of our research findings. First, the Flint water crisis resulted in criminal and civil lawsuits, which are still pending and increasing. Plaintiffs include residents, the State of Michigan, and the NAACP. Defendants include members of the state and local governments, hospitals, the media, and engineering firms. This makes it difficult to carry out qualitative research involving interviews or focus groups with key participants. Access to reporters and news organizations in Flint has proven challenging due to legal concerns. As result, litigation hindered our efforts to interview actors involved or affected by the crisis.

Evaluation of the media offers a one-sided representation of a crisis event; therefore, future research should incorporate residents, as did ^{Driedger, Mazur, and Mistry (2014)}, who documented the “evolution of blame” in the years following a water crisis in Walkerton, Ontario. Their work with focus groups found skepticism and distrust of government persisted, insofar as residents remained reluctant to drink what they admitted was probably safe water. So, in the future, similar focus groups and interviews should be conducted not only with residents, but also with government officials and other institutional actors.

Given the city of Flint’s demographics, future research should also include specific emphasis on how the media served its minority community and worked to include minority voices in its coverage. Previous research by Heider et al. (2005) finds minority communities want coverage on solutions versus continual emphasis on problems. Such research may be conducive to a longitudinal study of the Flint water crisis.

Despite the information about the quality of water during the crisis, the government repeatedly refused to respond or was unwilling to acknowledge errors in water testing. Transparency and open government functioning also failed to provide citizens with information about the hazardous conditions of their water.

The media also failed to search for other sources or doubt the veracity of government statements. This is a clear case of lack of accountability. The more informed a citizen is, the better he/she will be prepared and able to respond when facing a crisis.

The Flint case demonstrates not only a failure to communicate health risks, but also the results of an unaccountable government that offered “callous and dismissive responses to citizens who expressed concerns” ^{(FWATF, 2016: 2)}. Some of the harshest criticism centered on Michigan Department of Environmental Quality spokesman Brad Wurfel, who first referred to Dr. Hannah-Attisha’s comments as “near hysteria” and “unfortunate” ^{(FWATF, 2016: 8)}. Wurfel resigned in December 2015 ^{(Smith, 2015)}, and he now works with a Michigan law firm, where he is said to be advising clients on how to maximize their communication strategies ^{(Egan, 2016c)}. To date, Wurfel has escaped formal charges.

“Let me start here - anyone who is concerned about lead in the drinking water in Flint can relax.” - Brad Wurfel, spokesman for Michigan’s Department of Environmental Quality in an interview with Michigan Radio. July 13, 2015. ^{(Smith, 2015)}