GLOBAL CHALLENGES: EBOLA OUTBREAK "EVERYBODY'S BUSINESS": MOBILIZING CITIZENS DURING LIBERIA'S EBOLA OUTBREAK, 2014-2015 SYNOPSIS When Ebola crossed into Liberia in early 2014, the West African nation had few defenses. Because no effective vaccine was available at the time, the only way to limit the spread of the viral disease was to restrict physical contact with those who were infected, what they had touched, and the bodies of victims. But that advice countermanded the most basic of human instincts: to comfort a sick child, hug an ill relative, or shake hands with a friend or coworker. The challenge of changing human behavior was especially difficult because Liberia was still recovering from a long civil war. Public distrust of government, persistent rumors, linguistic diversity, and limited communication capacity hobbled efforts to send a clear public message and win citizens' cooperation. After top-down tactics-including forcible quarantines of whole communities-failed to stem the rate of infection, a small team of Liberian officials, supported by international partners, realized that effective steps to contain the disease would require active participation by citizens themselves. The officials engaged Liberians in developing an information campaign and recruited people throughout the country to visit their neighbors door-to-door, explain the steps people could take to protect themselves, and respond to questions. Although the complexity of the Ebola response and the volatility of the outbreak had made it hard to measure the success of the social mobilization effort in reducing new infections, an analysis of timing together with anecdotal evidence strongly suggested that the effort helped save lives and contributed to the disease's decline during the final months of 2014. Leon Schreiber drafted this case study based on interviews conducted in Monrovia, Liberia in April and May 2016, with guidance and additional information provided by Jennifer Widner and Beatrice Godefroy. Princeton University's Grand Challenges program supported the research and development of this case study, which is part of a series on public management challenges in the West African Ebola Outbreak response. INTRODUCTION In March 2014, when Ebola virus disease started to spread from Guinea to Liberia, health workers, governments, and aid organizations were caught off guard. For the first few months, the official response was to stress the danger-"Ebola Is Deadly!"-and to blast that message over loudspeakers, broadcast it on radio and television, and splash it on billboards. But by late July, sounding the alarm had not stopped the rate of new infections from rising. The campaign offered Liberians little guidance about what they should do besides be afraid. "People were going by car and screaming [Ebola messages] over loudspeakers," recalled Dutch Hamilton, a Liberian who worked with the United Nations International Children's Emergency Fund (UNICEF). "But it was not working." The messages urged people to take steps that were intrusive. Families were to identify sick relatives so that case management teams could send them to isolation centers for quarantine until diagnosis. They had to abandon traditional practices of washing or touching relatives who had died, and they had to allow outsiders to spray their houses with powerful chlorine disinfectants if someone with the disease had visited. Safety required emotionally wrenching changes in everyday behavior. The warnings bewildered Liberians, many of whom chose to ignore them altogether. Others bristled at the government's attempt to interfere in their daily lives. Some believed the campaign was a fiction perpetrated by the government. By early August the rising rates of infection and deaths of frontline medical workers had fueled panic and pinpointed the need to reach out to people in new ways to help stop the epidemic. There were signs that many Liberians were searching for reliable information. UNICEF, which had expertise in improving communication during disasters and which had launched information campaigns as soon as the first cases appeared, used its partner radio station Hott FM to get the word out. When the station released a hipco (the Liberian version of hip-hop) song, "Ebola Is Real," which offered steps communities could take to be safe, it scored a broadcast hit.1 But rumor was still rife, and many people continued putting themselves in harm's way. Responsibility for mobilizing citizens to participate in the fight against Ebola fell to Reverend John Sumo of the health ministry's health promotion division. In mid August, the government created a new Incident Management System (IMS) to lead the response to the epidemic, drawing on a model the US Centers for Disease Control and Prevention (CDC) recommended. Sumo was appointed to head the IMS's social mobilization committee, one of six task teams, or committees. Each of the six teams was co-chaired by a Liberian official and an external partner such as a representative of a UN organization or a nongovernmental organization (NGO) that actively provided services related to a committee's mandate. Based on its global expertise in running community engagement programs, UNICEF became Sumo's co-chair. THE CHALLENGE The purpose of social mobilization, sometimes called community engagement, generally boiled down to four things: (1) building and maintaining trust between communities and the frontline staff who assisted in the response, (2) securing the adoption of prevention and control measures through face-to-face dialogue about the reasons for new practices and adapting the measures to unusual circumstances when warranted, (3) reinforcing messages about actions communities could take to protect their members-and countering misinformation, and (4) reducing the stigma against people affected. When people began falling ill in March, April, and May 2014, responders gradually began to appreciate the difficulties the social mobilization tasks posed. For instance, the first discovery was that the same words could have different impacts from one place to the next. The International Federation of Red Cross and Red Crescent Societies (Red Cross) had had experience with earlier Ebola outbreaks in Uganda, where the disease appeared in remote areas and quickly burned itself out. There, the Uganda Red Cross had worked with Médecins Sans Frontières (MSF) to contain the disease. Amanda McClelland, a Senior Officer in the Red Cross's Emergency Health Unit, said the communications team had used the message "Ebola Kills!" Although the fear the slogan generated might stigmatize survivors, the words were straightforward, direct, and effective in triggering behavioral change. The benefits seemed to outweigh the negatives, and it was relatively easy to manage the fear and stigma because there weren't that many cases. With detection of the first cases of the disease in Liberia, the natural temptation was to reach for what had worked in the past. The team borrowed what had worked in Uganda. But in this new setting, where there was no knowledge of the disease, the phrases created a quandary. Villagers asked, "If Ebola is deadly and there is no cure, then why would I go to an Ebola treatment unit if I will die anyway?" "Those messages caused a lot of problems when moved to West Africa," McClelland reflected. There were not enough experts on the ground early enough to adapt quickly. "There were a number of lessons, but the main one was 'context matters.'" As the Red Cross began to alter its communication strategy, a second challenge emerged: It was very hard for the groups leading the response to control the message. Marietta Yekee, Liberian deputy country director of the Johns Hopkins Center for Communication Programs-based Health Communication Capacity Collaborative (HC3), said, "Different organizations were spreading messages based on [Ebola information] they found on the Internet." The scattered, ad hoc, and often misleading information and advice Liberians encountered from loudspeakers, billboards, and radio shows across the country fed the public's skepticism and confusion. Finding the right words and coordinating the message were not the only challenges. Sumo, a respected Christian clergyman experienced in mobilizing communities during immunization drives and campaigns against other diseases like malaria, anticipated additional hurdles in trying to win broad public cooperation on Ebola prevention measures. Among them were a weak communication infrastructure, public skepticism, and lack of coordination among the faith groups, traditional leaders, aid groups, and international organizations that sought to help as the outbreak expanded. Reaching people was hard. A 14-year civil war, which had ended a decade earlier, had taken a heavy toll on infrastructure. Liberia's road network was limited, and only about 6% of road surface was paved.2 During the rainy season of roughly May to October, many of the dirt roads were impassable, and some of the paved roads had vehicle-size potholes. Studies indicated that less than half of the population had cell phone subscriptions at the time, and cell tower coverage could potentially reach only a bit more than 60% of the population.3 There were 22 FM radio stations in the capital, Monrovia, as well as another 40 across the country, and radio was a potentially powerful conduit for Ebola prevention messages, but most of the stations initially opposed the government. Communicating across Liberia's 111,000 square kilometers (43,000 square miles) was also problematical because of linguistic diversity. Citizens spoke 30 languages, whereas public officials spoke mainly in English, and radio broadcasts did not mirror the country's full linguistic spectrum. Therefore, to reach citizens, officials had to secure the help of local leaders, who could provide entrée to frightened communities, speak in terms people could understand, and locate residents more rapidly than outsiders could. Distrust created an additional hurdle. An Afrobarometer survey conducted in 2013 found that almost half of Liberian citizens reported having to pay bribes to tax officials and that 63% said the government was not combating corruption effectively.4 In 2014, Transparency International's Corruption Perceptions Index ranked Liberia 94th out of 174 countries, an improvement over previous years but still evidence of public skepticism. John Alexander Nyahn Jr., who served as executive director of Liberian NGO Community Health Education and Social Services, said that because of the country's record of corruption, communities did not trust information coming from the government. A survey of 1,500 residents in the capital city of Monrovia conducted in late 2014 by researchers affiliated with the International Growth Centre, a United Kingdom-based research network, would eventually confirm what had become obvious months earlier: that a pervasive distrust of government was "associated with lower uptake of preventative measures and lower support for control policies," even after accounting for knowledge of Ebola, sociodemographics, and fear of the disease.5 Perceptions of corruption also lent credence to conspiracy theories. Some Liberians concluded that Ebola was a government ploy to attract funds from international donors, for example.6 They pointed to the country's heavy reliance on foreign aid-which accounted for 73% of Liberia's gross national income in 20117-and claimed that the disease was a desperate bid to maintain the flow of money. A final challenge to social mobilization arose from poor coordination among the groups struggling to contain Ebola's spread. NGOs, government offices, traditional leaders, and religious figures sometimes dispensed divergent advice and conflicting recommendations. Community confidence suffered when people followed instructions only to find that promised services such as rapid ambulance response or the provision of chlorine supplies were unavailable.8 Peter Harrington, communications adviser of the United Kingdom-based Tony Blair Africa Governance Initiative, noted that in the early months of the outbreak, the communication arm of the response "was making promises that the hard infrastructure couldn't keep." In mid August, frustrations boiled over when Ebola began sweeping through West Point, a poor and densely settled neighborhood in Monrovia. The hospital that served the area had closed two weeks earlier after health workers had succumbed to the disease. Bodies began to pile up in the streets. The government hastily converted a school building in the community into a temporary Ebola isolation center. And anger grew when infected people from outside West Point were brought to the facility, leading to fears the government was importing Ebola into the neighborhood. On August 18, West Point residents rioted and looted the Ebola isolation center of its supplies and equipment, including contaminated materials. With no system in place to locate the dozens of people potentially exposed to Ebola during the chaos, government forces cordoned off the entire community. The forced quarantine enraged the community, and in subsequent unrest, the military opened fire, killing a teenage boy and injuring two other people.9 The West Point incident exposed the weakness of early efforts to help people take steps to protect themselves and their families. FRAMING A RESPONSE Sumo's team had to develop a communication strategy adapted to a volatile and fluid situation. The team's end goal was to communicate crucial Ebola information effectively and to win compliance with steps that would reduce risks and end the epidemic. But to address the constraints the team faced, the team would have to find a way of devolving many decisions to the local level in order to enable trusted community members to take the lead while also ensuring that the information people received was correct and phrased in helpful ways. But those two ambitions could potentially conflict with each other. According to Patricia Omidian, an anthropologist who worked with the World Health Organization (WHO): "In the beginning, [Ebola communication] was 100% top-down-government edicts, basically. Traditional and community leaders were not involved in the early stages," she said. But that approach ignored the importance of linguistic diversity, poor infrastructure, distrust of the government, and the basic tenets of effective communication between sender and receiver. It became obvious that one-way messages carried on billboards, murals, radio, and television were failing to affect the behavior of many Liberians for at least two reasons. First, the sender (the government) was suspect. And second, signs and traditional media offered the receivers (the public) no opportunity to question and discuss a complex situation that had deeply personal ramifications. Two-way interaction with someone a person trusted could sharply improve the chance that the message would have an impact. Yekee with HC3 said: "If we were going to change the message content to focus on the fact that you can survive and to get people to believe us, we needed one-on-one discussion to explain the facts. We needed people to ask questions and talk about [Ebola]. The radio alone was not enough." The identities of the people tasked with initiating such conversations mattered. Nyahn, whose organization focused on improving health and sanitation in north-central Nimba County, cited the importance of the message carriers: "We could not use those same public servants who are corrupt, whom people don't trust, to give information." The good news was that Liberia already had a partial model for a community-driven approach. Since 2010, the health ministry's director of community health services Tomba Boima had worked with international NGO Global Communities on a Community-Led Total Sanitation initiative in two of the counties where trust in government and outsiders was especially low and where Ebola was taking a high toll: Lofa and Bong. Boima and Global Communities had helped create and train Natural Leader Networks by building teams of village activists or enthusiastic volunteers to serve as facilitators. Outfitted with T-shirts and picture guides, the leaders in the network could (1) dialogue with their neighbors in their own languages about ways to improve sanitation, (2) offer suggestions, (3) reach out to other communities, and (4) serve as official points of contact with the health ministry's district environmental health technicians. Good ideas could bubble up through that system, and important messages could get out quickly and inexpensively-through people communities trusted.10 The question was whether it was possible to adapt and expand the method to help contain Ebola. By August, Global Communities had already convened two community meetings in Lofa County to try to counter denial, reduce contact with sick people, and discourage secret burials of those who had died from the disease.11 To overcome Liberians' distrust of government organizations and international outsiders as well as cope with language differences and transportation barriers, planners had to identify and recruit trusted leaders of local communities in areas not already part of the Global Communities sanitation project. They also had to create a system to train, support, and monitor the work of those social mobilizers, or Natural Leaders, as they went door-to-door to spread the word about Ebola among their friends and neighbors. Devolving activity to the local level did not obviate the need to help government agencies and Ebola responders speak with one voice, however, nor did it reduce the importance of combating rumor. Sumo's team also had to address those challenges, but to do so, it had to bring a wide range of people together to deliberate about the best words to use and to find ways to spot incorrect advice. If the epidemic entered new phases, the messages would have to change, so the coordination system would have to remain in place until the end of the outbreak. But there was no immediate agreement on how to proceed. GETTING DOWN TO WORK A common metaphor among aid workers during the outbreak was that organizing the response was like trying to build a ship at sea. That was especially true of communication strategies, which developed through trial and error. The separate elements unfolded simultaneously, with most of the effort concentrated in September and October 2014, even as the Red Cross-Global Communities social mobilization efforts and the UNICEF messaging, which had started much earlier, continued their races to inform and persuade.12 Creating a system Because the IMS social mobilization committee had to begin work before its own structure was completely in place, Sumo adapted operations repeatedly while the new communication campaign took shape and moved forward. When the committee began to meet regularly in early September, it included people from all of the roughly 60 different responding organizations represented, and the result was predictable. "They all had extensive response experience, and they all believed their way was worthy," said Jana Telfer, associate director of communication science at the CDC. "And they were all talking at the same time." (The number of groups later rose as high as 78.) Telfer said that with dozens of organizations reporting directly to Sumo, Sumo's "span of control was beyond what management research shows is feasible." In late September, Telfer worked with Sumo to divide the committee into smaller groups. The move aimed not only to ease Sumo's workload but also to delegate responsibility and sharpen the committee's ability to work quickly and effectively. Building on the existing structure of the Liberian health ministry's health promotion division, Sumo and Telfer created working groups for the four most critical functions of the social mobilization committee: (1) message and materials development, which coordinated and standardized the content of Ebola-related communications; (2) training, which focused on equipping social mobilizers with skills for engaging directly with communities about Ebola prevention; (3) field mobilization and support, which provided ad hoc assistance and media help for the group; and (4) media liaison, which worked with the ministry of information to strengthen coverage of the Ebola crisis. With the help of the Africa Governance Initiative, Sumo and Telfer next drew up lists of responsibilities for the working groups. They extended the IMS's practice of sharing management responsibilities by assigning one international partner and one national NGO as co-conveners for each of the working groups. UNICEF became the co-convener for message and materials development; Restoring Basic Health Services, a US Agency for International Development (USAID) project managed by John Snow Inc., led training; WHO managed field mobilization and support; and the CDC focused on media liaison (figure 1). No international partner or Liberian NGO partner was permitted to participate in more than two working groups. Weekly gatherings of the full social mobilization committee became more fruitful after the four working groups began meeting in advance to agree on progress and obstacles. Telfer said the working-group meetings "brought order to the main meetings." She added: "It was still 60 different people in one room, but they weren't all talking at one time. There were presentations from each group, and decisions were made by the group." The working groups "had basically done their business and were doing reports from their working-group meetings and making recommendations to the full committee." Building collaboration In late August, just as Sumo's committee was beginning to take shape, the first steps to harmonize messages were getting under way. At that time, people saw no point in trying to protect themselves because almost "every single message was still negative," said WHO anthropologist Omidian. "The really important thing was to change that message and tell people, 'Here's what to do.'" But even announcements from the IMS inadvertently reinforced a sense of fatalism. IMS manager Tolbert Nyenswah opened press conferences with "how many people died today," recalled the CDC's Telfer. Anna Helland, country director of the Johns Hopkins HC3 program in Liberia, added that "the message 'Ebola Kills' was actually necessary, because you had people saying [that Ebola virus disease was] not real." But, she added, "You want the fear high, but you also want self-efficacy and knowledge of what to do to be high too." Both tone and content mattered. At first, when messages called for citizens to change their behavior in specific ways, there had been no effort to offer reasons or explain why in terms that meant something to people. Omidian said: "The message people got in the community was, 'What you do is wrong and bad.' And yet what they were doing was being loving family members. Everything people needed to do to stop Ebola went against the basics of what it meant to be a good human. . . . And here comes the government, saying, 'Trust us.'" In late September, the message-and-materials-development working group threw itself into high gear. The goal was to become the central clearinghouse for all Ebola-related messages. In addition to tailoring specific messages for billboards, posters, radio, and television, the group wanted to evaluate materials from all outside organizations prior to publication or release. An important early task was to identify the many organizations that had their own information campaigns. Sumo and UNICEF-the co-conveners of the social mobilization committee-worked with Liberia's 15 county health teams to organize and map volunteers and NGOs so they could determine who was working where, said Rania Ellesawi, UNICEF program manager of social mobilization during the Ebola response. The group decided to use the "healthy life" logo (figure 2)-which the health ministry had developed in 2010 to brand its communications-as a stamp of approval for all of the Ebola messages the committee endorsed. Sumo told partners: "the [healthy life] logo needed to appear on those messages. If not, the messages are not from us." According to Helland, "The main purpose was really to do no harm. You had so much happening that you just wanted to say, 'OK, your material could be better, but you're doing no harm with it, so go ahead.'" In practice, most messages got approved if they included no scientific inaccuracies and did not contradict other social mobilization messages. But the group also tried to identify which approaches worked best. Not surprisingly, given the diversity of aid organizations in Liberia at the time, the review process ran into opposition. Sumo recalled that "some [organizations] still did not want to go through the vetting process because they felt they could do it on their own." If an organization released messages without the committee's approval, Sumo reported the case to IMS management. "The IMS had a representative from the president's office, Emmanuel Dolo [who had a doctorate in social work]," he added. "We would report any issues to him, and as a representative of the president, he would then raise a strong alarm." Refining the message The message-and-materials working group next focused on strengthening message content. In some cases, partner organizations submitted proposals to the working group. The team sought a new approach "that was based in science, that switched the frame from negative to positive, and that included cultural considerations," Telfer said. The team worked closely with the epidemiological arm of the response effort to make sure that the information it provided about the things people could do to protect themselves was scientifically sound. But the team also focused on ways to pitch the information so that people would pay attention and take the advice to heart. "We're facing the same issues Coca-Cola faces. Different messages are more effective with different audiences, and different routes of delivering messages are more effective with different audiences," Sheldon Yett, UNICEF's country representative in Liberia, told The Atlantic.13 Helland emphasized that the process for testing and refining messages "was very quick and dirty" and took 36 hours or less. "People would submit a message today, and we would go into the community and do rapid surveys or hold focus groups. By the next day, the results were available," Yekee said. Through those focus groups, the working group learned the types of messages that appealed to Liberians. "People uniformly said they responded better to positive messages," Telfer said. That recognition was in line with research findings in other contexts.14 Further, testing showed that Liberians preferred language that stressed unity. "We had to consider very deliberately that the country was only a decade out of a devastating civil war," Telfer said. "Our messaging had to look at the cultural importance of talking about how 'United we could do this; together we could participate.'" To help organizations craft the advice they gave communities, the message-and-materials working group also created a manual containing approved Ebola messages. The manual enabled partner organizations and government agencies to release communications quickly during emergency situations. The first version of the manual, published in late October, was eight pages long and highlighted three primary-prevention messages. Two lists of supporting messages included information on the way Ebola is spread, on key contacts, on survival, on safe burials, and on community engagement.15 The manual also laid out the process for requesting message approval. As coordination improved and the outbreak evolved, the working group revised the manual, with the fifth and final version published a year later, in October 2015. The document ultimately grew to 26 pages and included messages on the restoration of routine health-care services. The tone shifted, too. The text spoke of "remaining vigilant in an Ebola-free Liberia," listed ways Ebola is not spread, and provided information on how to access the government's newly opened cemetery for "respectful" burials.16 Telfer emphasized that "Liberia has, to this day, the most sophisticated message manual of any of the three countries most affected by Ebola. If Ebola happens again, the country has a concrete message manual to serve as a reference." The message-and-materials-development group took another important step later, in November 2014, by reviewing and overhauling the primary national media campaigns, including "Ebola Is Deadly" and "Ebola Is Real," the two themes that had dominated the communication effort through October. Sumo said information from focus groups and GeoPoll, a mobile platform for conducting surveys, shaped the new campaign. (See results from the GeoPoll survey, administered by Johns Hopkins's HC3 program in November 2014, in text box 1.) Science, too, played a big role in shaping the message. Harrington of the Africa Governance Initiative said the message-and-materials working group approached the IMS's epidemiological committee to ask, "What are the five behaviors that if everyone in the country changed them would get us to zero?" The scientists identified "safe burials, don't touch, report cases, cooperate with contact tracers, and stay in quarantine," he said. Armed with survey data and the scientific information, a group of Liberian officials and international partners created the first draft of a new campaign. According to Harrington, the group "worked 15 hours a day over the weekend to hash out the draft and fit it on one page." The short format increased the willingness of senior decision makers to devote time to the effort. After everyone in the group reviewed the draft, Sumo presented a new campaign, "Ebola Must Go: It's Everybody's Business"-to the full social mobilization committee and won approval. Sumo emphasized the significance of the new campaign: "We were now graduating from 'Ebola Is Real' and 'Ebola Is Deadly and There's No Cure' to 'There Is Hope [if you go to a treatment center when you notice symptoms].'" President Ellen Johnson Sirleaf launched the campaign in early December.17 The message subsequently rolled out on billboards, television, and posters. In addition, it appeared in a chart-topping song on radio stations all across Liberia. (text box 2). Because of the country's linguistic diversity and an adult literacy rate of only 43%,18 the posters, flyers, and other materials created for the "Ebola Must Go" campaign appeared in 16 vernacular languages. The materials also conveyed messages through brightly colored illustrations more than through words. "To ensure that people interpreted the information accurately, illustrations clearly specified the nature of the symptoms [e.g., vomiting, diarrhea] and the action to take" [e.g., tell a community leader], Telfer said.19 Media liaison Journalists played a central role in conveying important information to Liberians across the country. However, to avoid confusion and error, which fed rumor, it was important that all parts of government offer the same message and that one ministry take the lead. Early in the crisis, some radio stations had broadcast misleading information, according to Liberian Adolphus Scott, who headed UNICEF's Communications Development Program during the Ebola outbreak. "[Radio stations] alleged that Ebola was a conspiracy so the government could get more donor funding, and communities listening to those messages were resistant," he said. Still, the country's 60 or so radio stations represented one of the few ways to reach the large numbers of Liberians who lived in remote areas. When the number of new Ebola cases escalated in July and August, before Sumo's new committee was in place, the health ministry arranged meetings with community stations "to get them on our side," Scott said. Coordination within government was part of the challenge. During the early months of the outbreak, the Ministry of Information, Culture, and Tourism-the Liberian government's official public communication agency-was responsible for arranging daily press updates on Ebola. In August, when the IMS took shape, it required efforts to define roles and develop working relationships between the two groups. At first, the daily IMS meeting did not end until 11 a.m., the moment at which the information ministry conducted its regular press briefing, broadcast live on radio. As a result, IMS manager Nyenswah had to race across town to the ministry to report on the latest Ebola death toll, and he arrived late every time. That timing problem undermined the goal of clear communication. In late September, the media liaison working group, co-led by Richard Zeon and Lahannah Jawara from the health ministry and Telfer from the CDC, worked to improve the relationship between the social mobilization committee and the information ministry. "We had to build a bridge between the IMS and the information ministry," Telfer recalled. She and Harrington, from the Africa Governance Initiative, worked with their Liberian colleagues to create a joint communication committee led by information minister Lewis Brown, who invited other government agencies to join the effort.One of the goals was to provide more-constructive and more-useful information for the ministry's daily press briefings. Telfer said the group decided to change the earlier approach of leading the briefing by announcing the numbers of people who had died. "We didn't delete how many people had died, but we talked about what we were doing to bridge the gap. We also began to give Nyenswah information about process and about what the IMS had done that day." The goal was to "provide the IMS manager an opportunity to evaluate and analyze his information and also to provide the country an opportunity to focus on something besides people dying." The media liaison group also worked with Nyenswah and Brown to revise the structure and timing of the news conferences. Instead of hosting daily press briefings, they shifted the conferences to Mondays, Wednesdays, and Fridays, rebranding them as "health focus" briefings. The less-frequent schedule provided the working group with opportunities to consider information as it was coming in and to be more deliberate about sharing." The social mobilization committee also introduced a two-page Ebola update bulletin that was sent to the information ministry and all international partners three times a week. The IMS upgraded the press briefings' perceived quality and credibility by including more participants who had direct knowledge of what was going on. In addition to Nyenswah, the leaders of individual IMS committees were given the opportunity to speak at the news conferences. The media liaison team also made sure there was an international partner at every briefing "to show that a lot of collaboration was happening," Telfer said. Later, in 2015, when the goal was to return the country to normal footing, the ministry used Tuesdays and Thursdays to invite other government agencies to report on their activities so that people could see that those government agencies were busy running schools and other services. Creating an army of Ebola foot soldiers Improved coordination, clearer messaging, and community radio were not enough to overcome the public's deep distrust of official sources. Scott of UNICEF said people realized "we sit in our offices, cook up messages, and take them to these traditional homes-but in most cases, we missed the boat." The measures did little to help win hearts and minds in the most-isolated areas, where language barriers, distrust, or low radio listenership interfered with communication. To address such challenges, it was crucial to create a role for local communities in building understanding and solving problems-and to then listen to what they said. Sumo's team worked with the health ministry to adapt and expand the Global Communities model of community engagement and involve other groups in similar efforts. USAID helped finance the creation of the community engagement program in late August and expanded its support in succeeding weeks. In the early phases of the outbreak, 600 Natural Leaders became social mobilizers or Ebola foot soldiers.20 There were certain basic things people had to know in order to protect themselves, such as how to use chlorine to prevent infection and what to do to protect oneself if an ill family member vomited. For more-complicated problems-for instance, what to do if a powerful member of the community violated the rules or if a neighbor's guest suddenly became ill-past experience had shown that face-to-face discussions, wherein people could ask about how to deal with specific situations, were more effective than handing out leaflets or than broadcasting instructions to large groups. Telfer said, "The [low] literacy level, poverty and poor infrastructure made face-to-face communication the best means to convey information."21 By early October, members of the social mobilization committee were developing projects to launch or intensify door-to-door information campaigns. But before they could undertake any such programs, the committee had to find ways of persuading Liberians-many of whom remained skeptical of the response-to welcome these so-called social mobilizers into their communities. The key to responding to local challenges was to "identify people's perceptions at an early stage," said Kabah Trawally of the National Mandingo Caucus of Liberia, an NGO that led seminars and other activities. Members of the social mobilization team concluded that to build relationships with affected communities, they had to "bring the chiefs and religious leaders on board," Sumo stressed. Permission from those influential people was a common requirement for strangers who wanted to visit village households. Without such endorsement, the team would have little ability to meet people where they lived. Just a few years earlier, the government had created a National Traditional Council, sometimes called the Council of Chiefs and Elders, to represent the hundreds of chiefs and elders across the country's 15 counties. In partnership with the United States-based Carter Center's Access to Justice in Liberia project, which had a long history of working with traditional leaders in Liberia, the social mobilization team organized a series of meetings with the council. The first meeting, which took place over several days in early October in Bong County, brought together every ministry and committee involved in the IMS in order to brief the traditional leaders and ask for their support in the Ebola fight. A second meeting followed in another county a few days later. Scott noted that with Ebola raging in communities throughout the country, traditional leaders "were very receptive." The council adopted a 10-point resolution to support the government's Ebola response. And with traditional and religious leaders on board, the committee could launch the door-to-door campaign. The health ministry and UNICEF ran one major door-to-door program for social mobilizers, which it called Reach Every District. The idea was that community health volunteers, a group of people previously selected by their communities to work with the health ministry, would offer practical information about infection prevention and would spot problems. The idea was to teach those volunteers to train some of their neighbors, who would then visit homes to speak with families. At the same time, international NGO Mercy Corps, with support from USAID, introduced a parallel effort that enlisted 77 nongovernmental groups, mainly Liberian, to go door-to-door. It called its campaign the Ebola Community Action Platform. Mercy Corps's program director for the Ebola response in Liberia, Catherine Brown, said, "If you think about it, you probably would never subcontract 77 organizations in an emergency, because how do you implement and monitor this? On the other hand, under this cloud of catastrophe, there really was no other way to reach so many people." Together those two programs came to account for the vast majority of the thousands of social mobilizers who went door-to-door throughout Liberia from October 2014 to May 2015. The message-and-materials-development working group designed a standardized, daylong training curriculum for both programs. Yekee explained that because the health volunteers the ministry recruited were already well versed in methods of community engagement, the new curriculum had a really practical focus: "It included things like how to use a plastic bag to stay safe if someone is vomiting at home and you wait for help to come." The training designed for Community Action Platform social mobilizers was more extensive because the community members the local NGOs recruited usually had no prior experience with health communication. Population Services International, a global health NGO, worked with Mercy Corps, to create that training, entitled "Listen, Learn, Act." The guide used pictures as well as words to help communicate because some of the volunteers could not read. The approach stressed the need to listen. Annie Singkouson, a technical adviser from Population Services International who helped design the material for the Ebola Community Action Platform, stressed that during the listen step, it was vital to allow community members to do most of the talking. "Rather than [the social mobilizers'] speaking and saying, 'This is what you should do,' you wait for the person to give you more information on what they believed." Follow-up questions were important parts of the process. The next step-the learn phase-was to correct any mistaken assumptions participants held, as well as to pass on other correct information. The final step-act-was to ask what actions the community could take to prevent Ebola (figure 3). Singkouson said the social mobilizers "were not to tell them what to do; communities were supposed to come up with their own solutions." The approach was predicated on the understanding that "because not every area was affected in the same way, we had to make sure the training material was flexible enough for partners to add their own local flavors to it." The health ministry program trained 20 staff members, including some UNICEF representatives, as master trainers and sent them to each of the 15 counties to train two additional staff members from the county health teams. In turn, the county officials trained four people in each of the country's 88 health districts. Those people then recruited and trained community volunteers. Sumo said that on average, the program trained "114 community volunteers at the district level, including 45 chiefs, for a total of more than 10,000 people." The local people did house-to-house visits. The Ebola Community Action Platform took a similar approach. At the national level, the program trained people from the 77 partner NGOs, who in turn recruited and trained social mobilizers from the communities where they worked, including teachers (schools were closed at the time) and religious leaders. The mobilizers helped coordinate the communicators, who went door-to door. According to Brown, the program eventually deployed 15,000 community members. To compensate the thousands of people involved in the effort, the health ministry introduced a standardized payment of $50 per month for both programs' door-to-door volunteers. To avoid the possibility of corruption, the payments were earmarked to go directly to people rather than to district health budgets. Those who conducted the training for each group became monitors alongside other health ministry personnel. Monitors had to follow up with each of their communities at least once a week. There were also weekly formal reports on the number of meetings that had taken place and the kinds of questions or problems that had arisen in the door-to-door visits. The Mercy Corps Community Action Platform also implemented spot checks and a performance-tracking system. The emergency context of the Ebola response required that both programs have informal, ad hoc monitoring arrangements to respond to unexpected events. District or county mobilizers could phone for help and had guidelines to follow for reporting to local authorities or others. The Community Action Platform used a WhatsApp cell phone messaging network for ad hoc monitoring and response, which enabled the program to enlist the advice of a team that Mercy Corps ran. Despite those efforts to monitor performance, however, Brown said, there were inevitable weaknesses. "If you imagine the scope of the program . . . with thousands of 'employees,' you're going to have some issues, but maybe because of the context of this particular emergency, it did work," she said. "But the most important thing was to devolve power and trust in community expertise. As a result, the monitoring was more about triage because realistically, you can't manage everyone in such an emergency." In a quarterly report dated January 2015, Mercy Corps said, "The design of E-CAP [Ebola Community Action Platform] as a sub-granting program is inherently challenging in an emergency response where the emphasis is typically on speed and closely-controlled implementation."22 To ensure transparency, Mercy Corps ran a competition to select grantees, and that process slowed program implementation. Furthermore, the diversity of NGO approaches, although necessary to cope with different community contexts, presented challenges in an emergency context in which close conformity to standardized messaging was necessary. And as a result, the program did not really swing into action until January. Adapting Door-to-door social mobilization made it possible to adapt to local circumstances. UNICEF social mobilizer Hamilton pointed out that that capacity was essential: "On the ground, people sometimes asked questions whose answers were not in any book." Nor was everyone equally predisposed to try to learn about the disease and the measures required to prevent it. For example, people in communities already directly affected by Ebola differed from people in areas where the outbreak was not widespread. "Generally speaking, in communities that had firsthand experience of Ebola, people more readily accepted information about the symptoms and the actions required. Where it was not so prevalent, social mobilizers focused a lot more on explaining that Ebola is real and what the disease looked like," said Singkouson. Another variation was the difference between devoutly religious communities and moresecular areas. Trawally of the Mandingo Caucus was extensively involved in social mobilization in the predominantly Muslim communities around Somalia Drive in Monrovia, where tailored, flexible solutions were also necessary."The imams in the area had a really hard time accepting the government's announcement, in early August, that all dead bodies in Monrovia had to be cremated. . . . They were one of the biggest sources of resistance," Trawally said, explaining that it was traditional practice to use the same casket to bury deceased members of the congregation and that they were unable to persuade the imams to change the practice until the grand mufti, the highest official of religious law for the community, "made a ruling that it is what you do on Earth that counts, not when you are dead. He also put it in the context of the Prophet's teachings, which helped a lot." On New Year's Day 2015, the government lifted the requirement for mandatory cremations but still insisted on the need for safe burials. According to Trawally, the congregations immediately reinstituted regular burial practices and did not allow the government's burial teams to conduct safe burials for members. "As a result, we came up with the idea of a Muslim safe-burial team," he said. His organization gathered more than 30 imams from the community, listened to their concerns, floated the idea of a Muslim burial team, and then asked the group to recommend people to serve, following a Listen, Learn, Act approach. The Muslim burial-team idea was quickly implemented, with the government immediately sending people to train volunteers on how to conduct safe burials. Trawally's team also took the imams on a tour of the government's newly opened cemetery in Monrovia. "We had to convince them there was a demarcated area that would be used to bury only Muslims," he said. "And we didn't call it an Ebola burial site; we called it a safe burial site." Based on his experiences in Nimba County, in the north of the county near the borders with Côte d'Ivoire and Guinea, Nyahn echoed the importance of religious differences. "In one community, we used a Muslim fellow to talk to a Christian community. It did not work, though, because the community did not trust him as a mobilizer." He explained that they became aware of the problem only after a member of the Christian community phoned a radio talk program to voice his discomfort in discussing personal matters with a person from another religion. "So we had to swap him," Nyahn said, also citing other sources of variation, including differences between urban and rural areas: "Very rural communities tended to be very traditional, so we had to get the elders on our side. Once you get them, you have community entry." The other local challenge that Nyahn and other mobilizers confronted in Nimba County was proximity to Guinea, where the Ebola outbreak had originated. "Nimba has a strong cultural similarity to Guinea. On the border, social and cultural interrelationships served as key barriers" to Ebola prevention, he explained. "We could not separate Nimba's communities from Guinean communities along the border." Nyahn's concern was that even if communities on the Liberian side of the border managed to eliminate Ebola, open borders meant that the virus could simply cross over again from Guinea. Nyahn's organization's first step was to work with the UN Mission for Ebola Emergency Response to develop an accurate map of the region. "We mapped out the nine official crossing points, the 32 most frequently used illegal crossing points, the health facilities, and where immigration staff were," Nyahn said, adding that population sizes of villages along the Guinean border ranged from 600 to 1,200 and that the "population was very mobile and continuously moved between borders." The mapping exercise led them to conclude that "even if you were to carry all of Liberia's military forces along that border, you couldn't stop people from crossing." Instead of using force, Nyahn said, his group decided to work with people in the border communities to watch for unexpected visitors. "We used community members themselves to serve on the community watch teams. We coached them to know that if someone went across the border and interacted with people there and came back displaying signs and symptoms of Ebola, this is what you need to do and these are the phone numbers to contact." The keys to responding to local challenges were (1) to identify what people knew about Ebola and how they perceived the risks and the actions they could take and (2) to design a local communications strategy accordingly. "We couldn't have a general strategy that could work for all," Nyahn said. The involvement of local community members was crucial. Responding to rumors In spite of the social mobilization committee's efforts to inform the public about the facts, conspiracy theories and rumors about Ebola remained challenges throughout the crisis. Lack of coordination during the worst of the outbreak from June to October 2014 had made it impossible for the social mobilization committee to respond systematically to rumors and spurious information. But with a more robust system in place, the team moved to tackle the problem. Working with Internews, an international nonprofit organization that supported journalists and media professionals around the world, UNICEF's Liberia head of innovations Lee Kironget took the lead in developing a way to track and respond quickly to rumors. The idea behind the tracker was to take real-time monitoring and response efforts to the next level, Kironget said.23 Andrei Sinioukov, who became UNICEF co-chair of the IMS social mobilization committee and its message-and-materials working group in 2015, said one of the main sticking points was the definition of a rumor. "The word rumor has a negative connotation; it's kind of sensational and usually untrue," he said. "Although we wanted to hear about that, we also wanted to hear the positive stuff so as to really get the pulse of the community and also capitalize on the positives." With funding from USAID through the Johns Hopkins HC3, UNICEF and Internews together developed two SMS (short-message-service) platforms that enabled any cell phone user to submit information to a central database. Carol Doe, Internews Liberia's SMS system manager, said, "We thought a two-way SMS system would work best for Liberia because many people had access to mobile phones." (The system did not require users to have smartphones.) The two platforms fed into the same database called the rumor bank. The first launched in November 2014 and was called U-Report. UNICEF partnered with the Federation of Liberian Youth to recruit so-called U-reporters from all 15 counties. Sinioukov said that although U-Report was helpful, it was somewhat limited in reach: "You basically got younger people, not representation across the spectrum. At the same time, U-Report was critical for reaching those younger populations, which often served as vocal agents of change." In March 2015, another SMS platform opened; it was called Dey Say, a Liberian phrase that connotes rumors. By June 2015, the U-Report had 51,000 users, according to UNICEF.24 Both systems used the same cell phone number: 8737. Sinioukov explained: "To access Dey Say, you had to text rumor to 8737. You then received an automatic reply saying, 'Thanks for contacting the Dey Say system. What rumor would you like to report?' The user then had the opportunity to text a rumor to us." Users received a thank-you note that confirmed the information would be forwarded to a higher authority. The system was the same for U-Report, except the process started with a user texting the word join. Doe said that in practice "Dey Say's rumors were more specifically about Ebola and came from a wider base of users." When the two systems were launched, more than a thousand rumor reports came in weekly; and every Thursday, Doe downloaded the database for analysis and aggregation, including geographic coordinates. "We compiled the rumors along with the facts, aggregating them by counties in a newsletter we sent to the ministry of health, media houses, the humanitarian community, and Internews's network of journalists," Doe said. "The newsletter went out every Friday." Some rumors required immediate follow-up. "If someone sent a message that an Ebola patient was lying on the road, that was a rumor we had to confirm immediately," Doe said. "We would then call the local county health team or the CDC to go and verify the report." He added: "We then also used Internews's network of 16 journalists in every county to double check." UNICEF's Ellesawi said the system revealed how different facts often got mixed to create dangerous Ebola rumors. "At the time when schools were reopening in February 2015 after being closed for six months, we heard through Dey Say that some parents had taken their children out of school because they feared that they would be injected with Ebola." The first part of the statement was true; the president had closed schools at the end of July and reopened them as the epidemic waned in mid February. The second part was not true, but it was easy to understand where the idea had come from. At the same time schools were reopening, WHO, UNICEF, and the CDC launched a nationwide measles vaccination campaign to combat the spike in measles cases that had occurred in the wake of the Ebola epidemic. Ebola vaccine trials were also getting under way. The result, Doe said, was a rumor that "the government was trying to infect children with Ebola by pretending it's a measles vaccine." Thanks to the SMS platforms, the social mobilization team was able to mount a quick response. The first issue in the Internews weekly newsletter, released on February 21, 2015, explicitly highlighted the need for "media outlets and journalists to draw a distinction between routine [measles] vaccinations and the Ebola vaccine trial and to make sure parents and families are informed about the difference."25 The message-and-materials-development working group also incorporated that information into the community work of the social mobilization teams. UNICEF's social mobilizers met with members of NGOs and district health teams, who then spoke with local leaders and parents. "We would debunk rumors in this way on a weekly basis," Ellesawi said. "That was one of the biggest things in the final stretch." OVERCOMING OBSTACLES By early November 2014, Liberia's Ebola epidemic was easing.26 The rate of new infections had actually started to fall in early October, but at the time, the data were so confusing that no one knew. Although the virus's initial wildfire expansion was largely contained, however, the disease continued to simmer in densely populated areas around Monrovia, with occasional outbreaks in other locations. The fear was that any one of those localized outbreaks could cause the epidemic to grow again. So, the social mobilizers' job was to eliminate new infections, and that task was not easy. People could become complacent, new infections could crop up where people were especially hard to reach, or the virus could take off in a community where the level of trust in any sort of outside authority was especially low. A new challenge also surfaced: how to persuade communities to accept residents who had fallen ill but recovered from the disease and were now returning from treatment centers. Getting to zero depended heavily on the art of persuasion. Typical was a story that unfolded just a week before Christmas 2014. A family from Gbarnga, a Liberian town of about 35,000 people, returned home from the clandestine burial of an Ebola victim. The sad trip had exposed the family's 17 adults and children to possible infection by the Ebola virus, which had claimed 3,384 lives by that point. Worse still, if any family members who developed symptoms came in contact with neighbors, there could be a new outbreak. When a county health team attempted to quarantine the entire household at an Ebola isolation facility, family members drove the health workers away with rocks and threats. Then Hamilton stepped in. A Liberian UNICEF social mobilization supervisor, Hamilton visited the family soon after the clash and won their cooperation by putting himself in the shoes of family members. Instead of demanding that all 17 people undergo immediate 21-day quarantine, he first asked how the parents would feed their children if they could not go to their fields or to market. Had they received food and medicine from the county health team? When they said no, he phoned the health team and demanded food, which arrived a short time later. The family's hostility began to soften. Hamilton listened and asked more questions. He discovered that the family's main reasons for objecting to the quarantine had little to do with fear of the nearby isolation center. Instead, family members worried that their chickens, ducks, and dogs would die during their three-week absence-and they just wanted to spend Christmas at home. Hamilton talked with neighbors, persuading one of them to take care of the family's livestock. Then he offered a compromise that would buy time: If the family spent one week at a quarantine facility and no one showed Ebola symptoms, they could go home in time for the holiday. Hamilton eventually convinced the family to stay under quarantine for the full 21 days after two members displayed symptoms and were diagnosed with Ebola virus disease; they survived. Although the social mobilization campaign had swung into action after the initial downturn in the rate of infection, it was never assumed that the drop in infection rates would continue. Improved communications still had a job to do. The Internews newsletter of February 27, 2015, drew attention to the second unexpected challenge that emerged as the tide was turning in Liberia's battle against Ebola. "According to the United Nations, there are now between 5,000 and 10,000 Ebola survivors throughout West Africa," the report said. "These survivors . . . often face stigma upon returning to their communities."27 Helland explained that "as survivors returned home, communities said no, you shouldn't be here, you should be dead because Ebola kills." And if the survivors had initially brought Ebola into a particular area, community members often expressed animosity. "Some survivors left their communities because they weren't viewed as survivors; they were viewed as the people who had brought Ebola to their communities," Ellesawi said. The message-and-materials-development committee began to train social mobilizers to respond to those concerns. Later versions of the message manual included recommendations encouraging communities to "welcome survivors and anyone whose life has been touched by Ebola back into the home and the community. Encourage and help them to participate in community activities."28 Drawing on the now well-established relationships with local leaders, the field mobilization working group also organized ceremonies to welcome survivors back into their communities. When survivors were scheduled to return, the social mobilization team worked closely with their colleagues from a separate IMS committee on psychosocial support to prepare communities. "The two teams would go to the communities together to facilitate the reintegration process," Ellesawi said. "It wasn't a one-off thing, because social mobilizers would engage with communities on a daily basis. In some cases, the survivors became part of the social mobilization team doing house-to-house visits and other community engagements." Although stigma against survivors presented a new challenge, Ellesawi pointed out that, just like prevention of Ebola, the problem was largely "a communication thing." If people knew the symptoms and could understand that survivors were not threats, "they were more comfortable." With functioning systems in place, the team was able to act quickly and effectively when the issue emerged as a fresh obstacle to the response. The increasing number of survivors presented an opportunity for improvement of care capacity and for confronting pessimism and rumors. Sumo said survivors represented a powerful tool for the social mobilization team. "We incorporated some survivors into our campaigns, and they were able to help us spread the message," he said. "People started seeing survivors with their own eyes. Based on that, people began to accept" that they could survive Ebola. Liberian Janice Cooper, who headed the IMS's psychosocial support committee, added that "because survivors were technically immune from Ebola and wouldn't get infected again, there was this idea to now use them in hotspots and in ETUs [Ebola treatment units]." ASSESSING RESULTS Liberia's Ebola social mobilization team aimed to reduce the rate of new person-to-person infections to zero. Achieving that goal would require persuading people to change their behavior in order to protect themselves and others-rarely an easy task. During June, July, and August 2014, when official messaging had stressed danger and rumors were rife, the number of new infections each week regularly exceeded the number of the week before. In September, laboratory-confirmed new infections peaked at more than 300 new cases a week, with many more probable cases. Just before the new people-oriented campaign began in October and November, the rate of infection began to drop and continued to do so. The number of new cases per week fell sharply, and then slowly diminished further. Although the campaign likely contributed to the decline, many Liberians appeared to have acted independently to take steps to protect themselves and their families. The IMS's social mobilization campaign may have sustained and expanded those measures, but by themselves they could not account for the epidemiological patterns observed. To assess the effectiveness of its efforts, the communications team ideally needed to know how many people had changed their behavior after conversations with social mobilizers or after hearing some of the approved messages-and how many did not. But those kinds of data were scarce in real emergencies. No one knew either the number of Liberians who had distrusted outside authority, had dismissed Ebola as a government concoction, or were simply slow to pay attention for various reasons or whether or how their views changed in response to seeing people fall ill around them or in response to hearing key messages. Although some survey data was available, it was hard to completely separate the effects of the campaign from directly observing the disease's effects and taking corresponding countermeasures. The results of a March 2015 UNICEF-sponsored, comprehensive national study of knowledge, attitudes, and practices (KAP) on Ebola in Liberia implied a link between the campaign and what people thought and did. Based on surveys that began in December 2014, the KAP study showed that "general awareness of Ebola [was] virtually universal" by that month.29 All respondents said they had heard of Ebola. The KAP study also revealed that "recognition of Ebola symptoms and key behaviors to prevent transmission was consistently high . . . Most [people] associated symptoms of Ebola with vomiting (84%), diarrhea (80%), and sudden fever (44%). Overall, 99% reported three or more correct forms of transmission."30 The study, which was conducted in six counties, found that knowledge still varied across the country, however, suggesting that by December, when the surveys took place, the campaign had been more effective in some areas than others-or that some communities took steps to inform themselves whereas others did not. For example, in Grand Gedeh County on Liberia's eastern border with Côte d'Ivoire, more than half of respondents said people "can prevent themselves from getting Ebola by avoiding mosquito bites"; 42% said Ebola could be prevented by bathing with salt and hot water. Only 11% of respondents in Grand Gedeh County said they would not touch the body of a sick family member who died from an unconfirmed cause of death.31 Nonetheless, by December 2014, more than 95% of respondents said they supported the core strategies for containing Ebola. Those strategies included isolation of symptomatic patients, quarantine of affected communities, referral to an Ebola treatment unit for treatment, contact tracing, and safe burials. The KAP study pointed out that "even if attitudes do not correspond perfectly with actual behavior, the nearly uniform acceptance of such measures suggests the emergence of community consensus around prevalent messaging campaigns, at least in principle."32 Moreover, almost all respondents (98%) reported altering at least one of their individual Ebola-related risk behaviors. The most common changing involved regular hand washing (83%), followed by 36% who said they avoided large gatherings and contact with people showing symptoms consistent with Ebola. Although the study did not reveal how many people were invited to attend burials, 24% of all respondents reported that to reduce their exposure, they had declined to attend at least one burial ceremony.33 A full 93% of respondents reported that radio remained their primary source of information on Ebola, followed by interpersonal communication (39%) and house-to-house visits by social mobilizers (36%). Focus group discussions with respondents "suggested radio reports, health visits and person-to-person interactions sometimes worked to mutually reinforce one other."34 A notable finding was that 45% of those surveyed reported that they had personally engaged with other members of their communities to discuss, demonstrate, or persuade them about Ebola prevention.35 The result represented a remarkable achievement in light of Liberians' deep skepticism toward the Ebola response during the early months of the outbreak. Clearly, many of those in the public felt empowered to become change agents on an issue of compelling public interest. Other studies buttressed those findings, although all were based on surveys conducted in limited areas. For example, no communities in the vicinity of Global Communities sanitation projects-the model for the outreach effort-saw new infections. Global Communities also reported that villages where it launched its outreach process were 17 times less likely to experience an Ebola infection than were similar areas where outreach had not worked.36 An International Growth Center-sponsored survey conducted in Monrovia during December 2014 found evidence that the door-to-door effort had had impact. "Citizens who experienced outreach are more likely to adopt preventative measures, support control policies, and trust the government," the authors reported. "These patterns are unlikely to be explained by confounding factors."37 Jana Telfer, CDC associate director of communication science, who joined the Ebola response in September 2014, said, "Community learning needs to be supported with knowledge about what the appropriate actions are." She concluded, "Conceivably, people could have done it on their own, but [without the efforts of the social mobilization team] my hypothesis is that it would have taken longer than it actually did." Of course, the social mobilization campaign was just one of several factors involved in changed attitudes and actions. Personal experience, too, was a powerful supporting factor, and the sheer scale of the death and destruction wrought by Ebola had to have an impact. The social mobilization team "had the best teacher on our side: death," observed Professor Hans Rosling, a Swedish statistician and professor of international health who spent three months in Liberia working on the Ebola response. Rosling used an example to explain: "If one person in a village dies, they hold a funeral. Then a few days later, some of the people who attended the funeral also die. Then everyone understood [that behaviors needed to change]; you didn't need to have a lesson." Qualitative interviews conducted as part of the KAP study echoed that point, stating that "many groups . . . [started] to take the threat of Ebola seriously after witnessing death and experiencing [the] fear that swept the country in 'Ebola time.'"38 Although reducing the rate of new infections was the objective and the main metric of success, it was also no small accomplishment to get the elements of a social mobilization campaign into place. One of the major achievements was institutionalization of the message-and-materials-development working group as the central clearinghouse for all Ebola-related communications. Despite the fact that the working group was "massively overloaded" due to the sheer amount of material submitted to it, Harrington concluded, "it was a huge step . . . that people acknowledged that they had to get approval from this [group] before they put materials out." Ellesawi pointed out that toward the end of the outbreak, the group essentially "developed all of the materials, tools, and messages that went into anything Ebola related, including radio and television announcements, posters, stickers, and everything else. It was completely centralized." In a country of 4.2 million people, figures from UNICEF reveal that the Reach Every District program had engaged more than a million people in community discussions and that more than half a million households received door-to-door visits from 10,000 social mobilizers.39 From October 2014 to July 2015, the Ebola Community Action Platform also reached 2.4 million people through 15,000 mobilizers that delivered messages on Ebola prevention.40 REFLECTIONS AND NEXT STEPS One of the keys to success was the involvement of both traditional and religious leaders from October 2014 through the end of the crisis. Adolphus Scott, who headed UNICEF's Communications Development Program during the Ebola outbreak, bemoaned the lengthy delay before those traditional leaders became engaged. "It was a big mistake on our part," he said. "We went ahead and left the traditional leaders behind, so we had to do a U-turn to reach out to them in October 2014. We should have started with the traditional leaders from the outset." Dutch Hamilton, a UNICEF social mobilizer who worked hard to win the cooperation of a distrustful family from Gbarnga, agreed that community involvement was decisive: "You don't say, 'The government sent me, so you have to listen to me,' because there was a gap between the government and community dwellers. Community engagement was the key. It was the bridge between the government and the people." It had taken several months to scale up the Global Communities door-to-door campaign, agree on messages, and develop systems for combating rumor. The question facing the international organizations involved in the response, as well as the Liberian government, was how to organize community engagement and coherent messaging more quickly the next time there was an infectious disease outbreak in a similar context. Innovations for Successful Societies makes its case studies and other publications available to all at no cost, under the guidelines of the Terms of Use listed below. The ISS Web repository is intended to serve as an idea bank, enabling practitioners and scholars to evaluate the pros and cons of different reform strategies and weigh the effects of context. ISS welcomes readers' feedback, including suggestions of additional topics and questions to be considered, corrections, and how case studies are being used: iss@princeton.edu. Terms of Use In downloading or otherwise employing this information, users indicate that: a. They understand that the materials downloaded from the website are protected under United States Copyright Law (Title 17, United States Code). This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/. b. They will use the material only for educational, scholarly, and other noncommercial purposes. c. They will not sell, transfer, assign, license, lease, or otherwise convey any portion of this information to any third party. Republication or display on a third party's website requires the express written permission of the Princeton University Innovations for Successful Societies program or the Princeton University Library. d. They understand that the quotes used in the case study reflect the interviewees' personal points of view. Although all efforts have been made to ensure the accuracy of the information collected, Princeton University does not warrant the accuracy, completeness, timeliness, or other characteristics of any material available online. e. They acknowledge that the content and/or format of the archive and the site may be revised, updated or otherwise modified from time to time. f. They accept that access to and use of the archive are at their own risk. They shall not hold Princeton University liable for any loss or damages resulting from the use of information in the archive. Princeton University assumes no liability for any errors or omissions with respect to the functioning of the archive. g. In all publications, presentations or other communications that incorporate or otherwise rely on information from this archive, they will acknowledge that such information was obtained through the Innovations for Successful Societies website. Our status (and that of any identified contributors) as the authors of material must always be acknowledged and a full credit given as follows: Author(s) or Editor(s) if listed, Full title, Year of publication, Innovations for Successful Societies, Princeton University, http://successfulsocieties.princeton.edu/ References 1 Uri Friedman, "How to Make a Hit Song About Ebola," The Atlantic online, August 25, 2014, accessed at http://www.theatlantic.com/international/archive/2014/08/how-to-make-a-hit-ebola-song/378980/. 2 See USAID blog with comparisons to the size of Tennessee at https://blog.usaid.gov/2014/05/liberias-road-miles-and-miles-to-recovery/ and The World Factbook at https://www.cia.gov/library/publications/the-world-factbook/fields/2085.html, both accessed June 28, 2016. 3 The Mobile Economy 2014, GSMA Intelligence, accessed July 10, 2016, http://www.gsmamobileeconomyafrica.com/GSMA_ME_SubSaharanAfrica_Web_Singles.pdf. 4 Samantha Richmond and Carmen Alpin, "Governments Falter in Fight to Curb Corruption," November 13, 2013, accessed at http://afrobarometer.org/sites/default/files/publications/Briefing%20paper/ab_r5_policypaperno4.pdf. 5 Lily Tsai, Robert Blair, and Benjamin Morse, "Patterns of trust and compliance in the fight against Ebola: Results from a population-based survey of Monrovia, Liberia," International Growth Centre Economic Impacts of Ebola Bulletin Three, April 2015, accessed May 28, 2016, http://www.theigc.org/wp-content/uploads/2015/05/Economic-impacts-of-Ebola-Bulletin-3.pdf. 6 Sara Jerving, "Why Liberians Thought Ebola was a Government Scam to Attract Western Aid," The Nation, September 16, 2014, accessed May 16, 2016, http://www.thenation.com/article/why-liberians-thought-ebola-was-government-scam-attract-western-aid/. 7 Blair Glencorse, "A decade of aid dependency in Liberia," Devex, August 19, 2013, accessed May 16, 2016, https://www.devex.com/news/a-decade-of-aid-dependence-in-liberia-81634. 8 Danny Gold, "Left to Die: Liberia's Ebola Victims Have Nowhere to Turn as Treatment Centers Overflow," Vice News, September 19, 2014, accessed May 23, 2016, https://news.vice.com/article/left-to-die-liberias-ebola-victims-have-nowhere-to-turn-as-treatment-centers-overflow. 9 "Liberian Government's Blunders Pile Up in the Grip of Ebola," Time, September 2, 2014, accessed January 20, 2016, http://time.com/3247089/liberia-west-point-quarantine-monrovia/. 10 Global Communities, "Global Communities' Approach to CLTS: Enhancing a Proven Methodology for Liberia," accessed August 5, 2016, at http://www.globalcommunities.org/clts-case-study-2015. 11 "Global Communities Liberia Team Conducts Community Outreach to Combat Spread of Ebola," accessed August 10, 2016, http://www.globalcommunities.org/node/37904 12 For more on UNICEF's involvement, see https://www.unicef.org/cbsc/files/EVD_soc_mob_response_infographic.pdf. 13 Uri Friedman, "How to Make a Hit Song About Ebola." 14 Lisa Cobb, "Ebola is Real: Using Theory to Develop Messaging in a Health Crisis," Ebola Communication Network, November 24, 2014, http://ebolacommunicationnetwork.org/ebola-is-real-using-theory-to-develop-messaging-in-a-health-crisis/. 15 Key Message Guidance Package for the EVD Outbreak in Liberia, Social Mobilization Sub-Committee, Version 1, October 27, 2014. 16 Key Message Guidance Package for the EVD Outbreak in Liberia, Social Mobilization Sub-Committee, Versions 2 to 5. 17 "'Ebola Must Go' - And So Must Prejudice Against Survivors," NPR, December 9, 2014, accessed January 18, 2016, http://www.npr.org/sections/goatsandsoda/2014/12/09/369382711/-ebola-must-go-and-so-must-prejudice-against-survivors. 18 "At a glance: Liberia," UNICEF, accessed July 3, 2016, http://www.unicef.org/infobycountry/liberia_statistics.html. 19 Jana L. Telfer, "Life and Death Communication: Lessons From the Ebola Epidemic," PRNews, May 23, 2016, accessed July 10, 2016, http://www.prnewsonline.com/subscription/2016/05/23/life-death-communication-lessons-ebola-epidemic/. 20 Global Communities, "Global Communities' Approach to CLTS." 21 Jana L. Telfer, "Life and Death Communication: Lessons From the Ebola Epidemic." 22 Source: Quarterly Report - September 2014-January 2015, Ebola Community Action Platform, Mercy Corps, accessed July 11, 2016, http://pdf.usaid.gov/pdf_docs/pa00k9gc.pdf. 23 Communication for Development: Responding to Ebola in Liberia, UNICEF, June 2015, accessed May 26, 2016, http://www.unicef.org/cbsc/files/Liberia_Case_Study_U-Report_%283%29.pdf. 24 Communication for Development: Responding to Ebola in Liberia, UNICEF. 25 "Information Saves Lives Media Newsletter #1," Internews Liberia, February 21, 2015, accessed May 26, 2016, https://Internews.org/sites/default/files/resources/Liberia_Reports/Liberia_Media_Newsletter_1.pdf. 26 "Ebola outbreak: MSF confirms case decline in Liberia," BBC News, November 7, 2014, accessed May 27, 2016, http://www.bbc.com/news/world-africa-29957338. 27 "Information Saves Lives Media Newsletter #2," Internews Liberia, February 27, 2015, accessed May 26, 2016; https://Internews.org/sites/default/files/resources/Liberia_Reports/Liberia_Media_Newsletter_2.pdf. 28 Key Message Guidance Package for the EVD Outbreak in Liberia, Social Mobilization Sub-Committee, Versions 3 to 5. 29 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 10, March 2015, accessed May 29, 2016; http://www.unicef.org/cbsc/files/KAP-Study-Liberia-March-2015.pdf. 30 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 10. 31 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 10. 32 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 11. 33 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 11. 34 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 11. 35 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 11. 36 Global Communities, Stopping Ebola in Its Tracks: A Community-Led Response, pages 41-43, http://www.globalcommunities.org/publications/stopping-ebola-in-its-tracks-2015.pdf 37 Lily Tsai, Robert Blair, and Benjamin Morse, "Patterns of trust and compliance in the fight against Ebola." 38 National Knowledge, Attitudes and Practices (KAP) Study on Ebola Virus Disease in Liberia, Ministry of Health and Social Welfare, page 10. 39 "UNICEF Liberia Social Mobilization Ebola Response," UNICEF, accessed July 10, 2016, http://www.unicef.org/cbsc/files/EVD_soc_mob_response_infographic.pdf. 40 "Engaging Communities to Prevent the Spread of Ebola in Liberia," Ebola Community Action Platform 2, accessed July 10, 2016, http://www.ecapliberia.org/images/ECAP2FactSheet.pdf. ISS is program of the Woodrow Wilson School of Public and International Affairs: successfulsocieties.princeton.edu. ISS invites readers to share feedback and information on how these cases are being used: iss@princeton.edu. (c) 2017, Trustees of Princeton University. This case study is made available under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Terms of use and citation format appear at the end of this document and at successfulsocieties.princeton.edu/about/terms-conditions.