Introduction

The newly emerged social media applications, with their participative, interactive, open, and transparent features, have enabled new ways of information transmission, interpersonal communication, and social participation (Campbell et al. 2014). This new technology is changing the landscape of the civil society sector all around the world by creating new ways of advocacy (Hefler et al. 2012), providing new strategies for mass mobilization online or offline (Gleason 2013), enabling the formation of new types of organizations (e.g., virtual communities, Komito 2011), and giving existing organizations new ways to disseminate information, build community, and mobilize resources (Guo and Saxton 2014; Lovejoy and Saxton 2012; Lovejoy et al. 2012; Waters and Jamal 2011; Waters and Lo 2012).

In China, too, the development of the civil society sector has been greatly influenced by the advancement in information technology, particularly the Internet (Tai 2006; Yang 2003; Yang and Calhoun 2007). Researchers showed that the Internet has facilitated individual expression and participation, which led to the formation of new nongovernmental organizations (NGOs). However, how the NGOs subsequently use the Internet to facilitate the achievement of their missions has seldom been studied. When it comes to the social media, partly because of the novelty of such applications, and partly because of the inaccessibility of popular social media sites in China (e.g., Facebook, Twitter and YouTube are all blocked by the Chinese government), there is virtually no systematic knowledge regarding the Chinese NGOs’ behavior on these sites.

Even though there are a lot of studies on the behavior of individual net-users (hereafter Netizens) and government censorship, the NGOs, as organizational actors, are in many ways different from human individuals (Geser 1992). For example, the interactions between human individuals are constantly molded by factors such as subjective cognitive impressions and emotional reaction which the individuals may not fully control (Goffmann 1963). Organizations, however, do not anchor their actions in psychological emotions. Instead, organizational activities are based on goals, values, or traditions (Weber 1972/1978). Thus organizations are “moral actors” in a more radical sense than human individuals. They are expected to take full responsibility for all actions they commit or omit (Geser 1992). In view of these differences between individuals and organization, we believe that the NGOs’ behavior online, their interaction with the government, the other organizations, and the general public may also differ from that of individual Netizens. For instance, an individual Netizen may feel it is safe to do something radical online, because he/she is anonymous online. An NGO, however, cannot be anonymous. This individual Netizen if caught by the authorities may argue that he/she is ignorant of the regulations or is under bad influences of peers. An NGO, however, may not claim to be ignorant of the law or to be a victim of deviant socialization.

Studies on NGOs and social media in other countries may help illuminate the Chinese case. However, since the Chinese NGOs, Chinese Netizens, and Chinese cyberspace are all very different from that of the West, it is unknown whether findings of the existing studies could be directly applied to China. For example, the Chinese cyberspace is heavily censored by the government, but the Chinese Netizens are found to be more active online than their Western counterparts (Gao et al. 2012; Sullivan 2012). Furthermore, as most of the Chinese NGOs emerged after the Market Reform in the new millennium, they are, in a sense, the digital natives, whereas their Western counterparts, with a longer organizational history, are likely to be digital immigrants. While the large nonprofits in the U.S. and Europe are learning how to adopt new social media technology, grassroots NGOs in China, particularly those formed online, might find it natural to use social media as the primary means of stakeholder engagement.

In this study, we adapted the Information-Community-Action framework (Lovejoy and Saxton 2012), which is developed based U.S. nonprofits’ behavior on Twitter. Using the framework, we explored Chinese NGOs’ behavior on social media by studying a sample of rural education NGOs on Sina-Weibo, one of the largest Chinese social media sites. By collecting and analyzing the posts of 155 Weibo accounts of NGOs for a period of 6 months (July 1st 2014–January 15th 2015), we have been able to describe the general patterns of social media use among rural education NGOs in China, focusing on what they say on Weibo, and how they say it. Then, we discussed factors that helped shape the NGOs’ behavior on Weibo. Finally, based on the findings, we proposed ways to improve the original Information-Community-Action framework, so that NGOs’ behavior on social media could be understood beyond the content level.

Information, Community, and Action: NGOs on Social Media

For any kind of organizations, in order to survive, they have to communicate to their stakeholders that their actions are desirable, proper, or appropriate according to certain socially constructed system of norms, values, beliefs, and definitions (Dowling and Pfeffer 1975; Suchman 1995). Through information disclosure, solicitation of stakeholder input, and responding to stakeholders’ needs and concerns, organizations gain legitimacy. Only then, can they mobilize resources for further development (Balser and McClusky 2005; Rupp et al. 2014). The advancement in information technology is providing organizations with fast and affordable ways to disclose information, maintain stakeholder relationships, and mobilize resources. The Web 1.0 technology, for example, enabled organizations to share information on performance, boost stakeholder responsiveness, and generate social capital via organizational websites and emails (Nah 2009; Saxton et al. 2007). Although the communication is largely one way (i.e., from the organization to the stakeholders), it is inexpensive, faster, and with a broader reach. Then, Web 2.0 technology brought about interactive elements, such as the Bulletin Board System (BBS), or a comment and reply section, which could be added to the traditional websites. These features could greatly facilitate two-way information flow, which is essential in establishing a sense of accountability and transparency in stakeholder relationships (Ospina et al. 2002). However, in the beginning of the Web 2.0 era, whether an organization can build and maintain an interactive website depends on whether they have staff members with the know-how. Some researchers proposed that a lack of IT-savvy staff could be one of the reasons why, in the beginning of the Web 2.0 era, NGOs have largely failed to use their websites as interactive and dialogic tools to engage their stakeholders (Kent et al. 2003; Saxton et al. 2007).

Now, social media applications such as Facebook, Twitter, YouTube, and Instagram provide easy-to-use interactive platforms, which require minimum IT skills. Any organization can use these applications at low or even no cost for information disclosure, stakeholder engagement, and organization promotion (Guo and Saxton 2014; Lovejoy et al. 2012; Waters and Jamal 2011; Waters and Lo 2012). They can make donation appeals, recruitment calls, donor acknowledgements, and public awareness campaigns online. In the process, organizations can also engage in conversations with “free agents” (e.g., bloggers, Twitter and Facebook users), which may make the organizations aware of new possibilities, including new funding opportunities and new programs (Kanter and Fine 2010).

To understand how social media is helping organizations with information disclosure, stakeholder engagement, and resource mobilization, Lovejoy and Saxton (2012) analyzed the content of Twitter posts of the 73 largest non-profit organizations in the U.S. and proposed a three-step hierarchical framework: Information-Community-Action. Lover and Saxton (2012) argued that social media applications have two broad organizational functions: information and dialog. Organizations first post information on social media to attract followers; then they use community-focused dialogic messages to bond with followers; finally, they use action-oriented posts to mobilize resources from the communities they have built. According to Lovejoy and Saxton (2012), generally, despite the ready-to-use interactive features of Twitter, the majority of the large non-profits tend to use it as a one-way information dissemination channel. Other studies on small community-based human service organizations, sport-for-development groups, and environmental advocacy groups had similar findings; organizations used social media primarily to market organizational activities rather than engaging in two-way communication with stakeholders and the general public (Bortree and Seltzer 2009; Campbell et al. 2014; Svensson et al. 2014).

A few researchers have attempted to explain such a phenomenon. Findings suggested that organizations’ adoption and usage of social media is closely related to organizational strategies (e.g., donor-focused, lobbying, or service-focused), capacities (e.g., size, pre-existing website, and public relation staff), governance features (e.g., membership structure, board size, and efficiency), and external pressures (e.g., funding structure) (Curtis et al. 2010; Nah and Saxton 2013). Service provision organizations, for example, are more likely to engage their customers through social media; while lobbying is found to be negatively associated with social media usage in the U.S. sample (Nah and Saxton 2013). Organizations with sufficient financial and human resources to set up an independent public-relation department are more likely to be active on social media (Curtis et al. 2010), whereas organizations with relatively fewer resources may open a free social media account, but fail to maintain it, let alone fully utilizing the dialogic communication features (Kim et al. 2014). Some scholars pointed out that the issue of social media adoption and utilization goes deeper than organizational strategies and capacities. They proposed that social media is not just a technology that one adopts, rather, it is a participatory culture or an attitude with an underlying philosophy of relinquishing control (Jenkins 2006; MacManus 2005; Macnamara 2010; Merholz 2005). In symmetrical dialogical communication, the organization can no longer control what followers do with the message on the social media platforms. Even though some studies showed that when two-way symmetrical communication is used the public demonstrates high levels of engagement with organizational messages (Bortree and Seltzer 2009; Cho et al. 2014), organizations may still shy away from adopting such communication strategies for fear of losing control (Quinton and Fennemore 2013; Waters and Feneley 2013).

Chinese Cyberspace, Netizens, and NGOs

Unlike their counterparts in the West who have had to adapt to the Web 1.0 era and then the Web 2.0 era, most of the Chinese NGOs emerged in the digital age. Market Reforms in the 1980s and the introduction of Western civil society ideas in the 1990s certainly contributed to the associational revolution in China (Lu 2009; Ma 2006; Zhang 1996). The Internet also helped the NGO-boom in China in several ways. First, with the help of the Internet, people who could not take advantage of conventional media now has a channel to express their opinions and to participate in the discussion of public issues and these in turn foster civic participation, including forming and joining organizations (Bach and Stark 2004; Shah et al. 2005; Yang 2003). Second, the Internet provides a low-cost means of communication, which enables people with similar interests/concerns to come together to organize regardless of their geographic location (Tai 2006). This can also facilitate the formation of ties across traditional socioeconomic boundaries and power differentials (Wittig and Schmitz 1996). Third, the Internet provides an easily accessible information platform. Hence, some events, when publicized online, will attract the public’s attention within a short time. Individuals may organize to respond to these events. Sometimes, these groups will dissolve after the problem is solved, or the situation becomes less severe. Sometimes, however, the groups will develop into formal organizations (Tai 2006; Zhou 2011). Last, the Internet lowers the transaction cost for the NGOs (Bach and Stark 2004). For some people that wish to establish an organization, raising donations online, recruiting members online, and promoting their organization online is perhaps the only affordable way. Furthermore, due to the unfriendly NGO registration and regulation system, many NGOs in China, particularly the grassroots groups, cannot obtain legal non-profit status (Kang and Heng 2008; Lu 2009; Ma 2006; Watson 2008). Setting up a website, a blog or a BBS is the only way they can show a somewhat legitimate presence to the public (Tai 2006; Zhou 2011). Considering the above factors, the Chinese NGOs’ attitudes towards the Internet and their behavior online are likely to be different from that of their Western counterparts.

Chinese cyberspace is also very different from that of the West. Many foreign websites, including popular social media platforms such as Facebook, Twitter, and YouTube are blocked by the “Great Fire Wall.” Domestic websites and social media applications are heavily censored by the government, thus greatly limiting what information individual Netizens are able to obtain and what opinions they can express online (Kalathil and Boas 2003; Li 2010; MacKinnon 2008; Xu et al. 2011; Yang and Liu 2014). Information considered offensive by the government is quickly deleted and individuals interested in discussing sensitive issues opt to bypass the Great Fire Wall and go on Facebook or TwitterFootnote 1. Some researchers have suggested that such a media environment discourages NGOs from using the Internet for social mobilization (Yang and Taylor 2010). This might be true to some extent. However, it is also important to point out that the political environment in general discourages Chinese NGOs from engaging in large scale social mobilization online or offline. In order to survive, many NGOs have to practice self-censorship to avoid offending the government, or sometimes trade organizational autonomy for government support (Lu 2009; Ru 2004).

Some researchers, however, have pointed out that, despite the control, Chinese Netizens enjoy a great degree of freedom online because, (1) the anonymous nature of the Internet makes it relatively safe, (2) websites, in order to keep customers, also find ways to resist government censorship, (3) many boundaries, rules and norms are still being negotiated and, (4) there are so many Netizens in China, and safety and power comes with greater numbers (Esarey and Xiao 2008; Herold 2008; Rosen 2010; Tang and Yang 2011). Although Chinese social media is dominated by entertainment news, gossip, commercial information and sports, dissatisfaction with government performance, or criticism of government policy is also prevalent in both explicit and oblique forms (Tang and Bhattacharya 2011; Tang and Yang 2011). Compared to their counterparts in other countries, Netizens in China are heavier contributors to social media in that they are more active initiators and commenters. They regularly make their views known and react to other people’s views (Sullivan 2012). Considering such distinct characteristics of the Chinese Netizens, we have reason to believe that the Chinese NGOs may also be more active than their Western counterparts.

Research on Chinese NGOs’ behavior on domestic social media sites is limited. Some researchers have attempted to study Chinese NGOs’ behavior on Facebook (Lo and Waters 2012; Waters and Lo 2012). However, the sample is likely to be biased, because it is unlikely for a common Chinese NGO to take the trouble to bypass the Great Fire Wall to maintain a Facebook page, unless its stakeholders are overseas. Furthermore, the domestic impact of these NGOs’ messages on Facebook is likely to be negligible when compared with those on popular domestic social media sites. For instance, studies found that Chinese Twitter users are more radical, liberal, and pro-democratic than Netizens on domestic social media. However, they are a rather marginalized group because online and offline control prevents their ideas from getting to the general public in China (Sullivan 2012). Likewise, even if a Chinese NGO can say whatever it likes on Facebook, it is speaking into a microphone with only a few listeners (Murthy 2011). After all, the domestic social media is so popular that even the government has discovered its power and is using social media to enhance its claims to legitimacy, circumscribe dissent and identify misconduct by its agents, as well as to adapt policy and direct propaganda efforts based on online public opinion (Sullivan 2014). Thus, it is important we study Chinese NGOs on domestic social media platforms.

To fill the knowledge gap regarding Chinese NGOs’ behavior on social media, the current study has explored posts that NGOs have made on Sina-Weibo, the biggest microblogging platform in China. We adopted the “Information-Community-Action” framework developed by Lovejoy and Saxton (2012). This framework classifies the posts based on functions offered by the social media platforms. Even though the Chinese cyberspace is heavily censored by the government, and the Chinese NGOs may have a unique relationship with the government, the Chinese social media sites still offer similar functions (i.e., information and dialog) as their Western counterparts (Shaoyong et al. 2011). Thus, the general categories of this framework (i.e., information, community, and action) are still applicable to the Chinese case. Some other researchers have used the “Information-Community-Action” framework to study NGOs and public institutions (Auger 2013; Guo and Saxton 2014; Neiger et al. 2013). In these studies, researchers kept the three general categories, but used bottom-up strategies to develop subcategories so as to better reflect actual organizational behavior in their specific social context. The current study followed the same strategy (described in detail in the next section). This not only allows us to describe the general patterns of social media use among Chinese NGOs and discuss factors shaping the organizations’ behavior, but also gives us an opportunity to compare our findings with findings on NGOs in other countries. We gain a better understanding of NGOs’ online behavior under an authoritarian regime.

Methods

Sample

The sample came from a previous study that involved an exhaustive snowball sample of NGOs working in rural education promotion in China. The original snowball process identified 464 NGOs (as of February 2011). To include newly emerged organizations and exclude organizations that have ceased to operate, the snowball process was repeated in June 2014, and a total of 490 rural education NGOs were identified. As we are only interested in the indigenous Chinese NGOs, organizations registered outside mainland China (e.g., registered in a foreign country, Hong Kong, Macau, or Taiwan) are excluded from the sample.

We chose to focus on rural education NGOs for three reasons. First, previous studies found that service provision organizations are more likely to be active on social media (Curtis et al. 2010; Nah and Saxton 2013). In our explorative stage, we would like to focus on organizations that are relatively active, so as to collect sufficient data to adapt the “Information-Community-Action” framework. The rural education field, which involves a lot of direct service provision (Zhou 2013), could serve as an example for service provision organizations. Second, rural education is a field that has received cross-sectoral support; the education promotion work often involves collaborative efforts from the NGOs, the local government, the schools, and many other parties. Thus, the organizations must handle multiple stakeholders from different sectors. Observing the rural education NGOs will allow us to see how NGOs interact with stakeholders from different sectors. Third, compared to NGOs working in sensitive fields (e.g., labor rights, and HIV prevention), rural education NGOs are less likely to post things that might offend the government. Thus, they were less likely to get their posts deleted or their accounts suspended during our data collection period. This will allow us to collect data with minimum interruption.

The Social Media Platform: Sina-Weibo

The social media we chose to study is Sina-Weibo (hereafter, Weibo), which is one of the most popular social media sites in China. Similar to Twitter’s interface, Weibo has a 140 character limit on the message that one publishes each time. These messages, called “posts,” are broadcasted in real-time and become part of the public stream. By September 30th 2014, Weibo’s monthly active user number had reached 167 million, and it occupied 87 % of the domestic microblogging market.Footnote 2 Business, NGOs, as well as government at all levels are using Weibo.

Data Collection

To identify organizations’ Weibo accounts, we first used Weibo’s search function to search for the names (or the abbreviated names) of the 490 organizations. Most of the organizations’ accounts were identified through this process. Some organizations, for various reasons, did not register their Weibo accounts using their names. A few provided a link to Weibo on their homepages. Others were identified by searching on Google.

Among the 490 rural education NGOs, 311 organizations have Weibo accounts. However, 156 organizations made fewer than 20 posts during the data collection period (July 1st 2014–January 15th 2015). These organizations were deemed inactive and dropped from the final analysis. In the end, 155 active organizations were included in the analysis.

We choose to observe the organizations from July 2014 to January 2015, because according to previous experience, the rural education NGOs, especially the grassroots ones, have their “peak seasons” and “quiet seasons.” They may be more active in the summer, because most of the volunteer teaching programs happen in the summer, and they need to raise money for scholarships before the fall semester starts. Then they may have a relatively quiet period of time, but get busy again if they need to organize in-kind donations for the winter, or raise money to cover the students’ expenses for spring semester. The chosen observation period covered both the “peak seasons” and “quiet seasons.”

All organizational posts published during this period were downloaded. Unlike Twitter, Weibo imposes strict restrictions on its application programming interface and encrypts all information published by its clients, making it hard to download the posts with software. Hence, all posts were downloaded manually. Seven student helpers were instructed to check their assigned accounts daily to copy the posts. In this way, we could also download the posts in a timely manner, in case some of them were removed by Weibo administrators later. The final database contained 34,145 posts. A graduate research assistant checked a random sample of 15 organizations to ensure reliability.

Coding

We used both bottom-up and top-down coding. We borrowed the Information-Community-Action framework proposed by Lovejoy and Saxton (2012), but developed the sub-categories within the three large groups based on a review of 5276 posts from the 490 rural education NGOs made in July 2014. A total of 16 codes were developed in this process (Table 2); the majority are similar to that of the Lovejoy and Saxton framework (with 12 codes). We developed four new codes to replace Lovejoy and Saxton’s single code “information,” and we added “Communicating with other organizations” in the community category.

The research team discussed the discrepancies between the codes during a double coding process (669 posts from 14 organizations made in July 2014) until 100 % agreement was reached. A coding manual was developed, specifying coding rules, giving examples for each code, and documenting our decisions regarding ambiguous posts. Using this coding manual, the rest of the posts were divided into three parts and coded by the research assistant and two student helpers independently. Each post was assigned a single code from Table 2. When dual or triple purposes appeared, we coded for the primary purpose. When it was impossible to distinguish between the primary and the secondary purpose of a post, we followed the “stronger” principle, i.e., to choose the more interactive code (action > community > information), because interactivity is the primary feature of social media. We randomly selected 341 posts to be triple coded. Krippendorff’s alpha (Kalpha) was calculated with bootstrapping. The triple coding has a Kalpha of 0.8619 (with only 0.0341 probability to fall below 0.80), indicating good intercoder reliability (Hayes and Krippendorff 2007).

Results

Overview of the NGOs in the Sample

The 155 organizations are from 28 different provinces/autonomous regions/municipalities. Beijing, as the country’s capital as well as its social, economic and cultural center, has 37 organizations. Guangdong, as one of the richest provinces, and home of Shenzhen, the pioneer in economic reform, has 23 organizations. The majority of the organizations are quite young, and were established between 2000 and 2010 (Table 1). Seventy-seven organizations (49.7 %) are registered with the government as non-profits. The rest have various identities, such as businesses (n = 5), affiliated entities (n = 6), and internal organizations (n = 45). Sixteen organizations (10.3 %) do not have any legal identityFootnote 3. Compared with other studies on rural education NGOs in China, the demographics of the sample is similar (Zhou 2013).

Table 1 Descriptive statistics of organizational variables
Full size table

In addition to Weibo, among the 155 organizations, 115 have traditional websites, 35 have BBS, 98 have blogs, 56 have Renren, and 52 have other types of online presence such as Baidu Tieba, QQzone, and Douban Group.Footnote 4

General Patterns of Weibo Usage

Many of the organizations in the sample have had their Weibo accounts for between 40 and 60 months. The number of followers, as well as the number of accounts the organizations follow, varies a lot. During our observation period, on average, an organization sent out 220 posts, with some publishing over 2000, while some sent out fewer than 30 (Table 1).

As shown in Table 2, 16 codes were developed to categorize the posts. They were grouped into the three functional categories: Information, Community, and Action. Among the three, informational posts were the absolute majority.

Table 2 General patterns of Weibo usage
Full size table

Organizational variation in the emphasis of these three categories is shown in Fig. 1. In this ternary plot, each organization’s placement is determined by the proportion of the organization’s posts that are informational, community building, and action oriented. As shown in the plot, the vast majority of the posts are informational. Generally speaking, there is a concentration of organizations in the right side of the tri-plot, which means a lot of the organizations are blogging to build community rather than calling for action. A few organizations are located in the middle, suggesting that they have a more balanced use of the three functions of social media.

Fig. 1
figure 1

Ternary plot: proportion of organizations’ Weibo posts in each category

Full size image

The Information-Community-Action Ladder

The majority of the posts we collected are closely related to the organization or the field it is working in. These posts fit well into the Information-Community-Action framework, and reflect the organizational legitimacy, accountability and resource mobilization theories: The organization shares information to attract followers, interact with the followers to build a community, and mobilize the community to take action.

Information: Advertisement, Voluntary Disclosure, and Mission Promotion

The Chinese NGOs like their Western counterparts spend most of their time on social media sharing information (63.7 % of the entire sample). The majority of the informational posts (33.7 % of the entire sample) were used to show off the organizations’ achievement, update the public about their most recent activities, and disclose financial and program evaluation information. Some of these posts are like advertisements, seeking to impress the stakeholders and the general public. For instance, Amity Foundation has become a U.N. consultant. The organization considered this a milestone in their development. They proudly updated their followers on Weibo in real time, saying: “This means Amity Foundation will have more opportunities to work with international NGOs, and that the voice of Chinese NGOs will be heard by more people.” Most of the other posts seem to be voluntary disclosures of the organizations’ performance in service provision, and financial matters. For example, this organization almost exhausted the 140 characters in describing a field inspection they had conducted:

School lunch program – field inspection in Guizhou: There were two dishes and soup for lunch today, sufficient in quantity and tasted good. Hygiene in the kitchen was OK, but we still reminded the staff to put the utensils in order and clean up the dust. The meal was allocated based on students’ needs. The dining hall staff kept detailed record for their work. However, when we compared the shopping record to the food we had, we found about 5 kilos of meat missing.

This is hardly a formal evaluation. In fact, systematic program evaluation is rare among Chinese NGOs, particularly the grassroots ones (Zhou 2015). However, it seems that organizations are still trying to show the public that they are concerned about program fidelity and impact, and that their volunteers and staff members are constantly trying to improve their own knowledge and skills.

Very often, organizations will attach photos and videos to their posts, or provide a link to a blog post or webpage, which allows longer text and richer information. For example:

The financial report of Cihong for August 2014 is published on Cihong’s website: http://t.cn/Rh4dUop, please check.

Photo sharing: Wang Bo Care Project – Scholarship award ceremony of the Mr. Wei Scholarship for needy students in Xinjiang.

Previous studies have found that vivid posts with videos and pictures are more popular on social media, and can attract more followers (de Vries et al. 2012). Moreover, among the Chinese Netizens, there is a saying, “With a picture, there is truth (you tu you zhenxiang).” Thus, it is not surprising that organizations have used pictures and videos to support their claims made in the Weibo posts. In fact, some of the posts we collected contained pictures only (or pictures and a simple emoji). It seems that the organizations let pictures speak for themselves.Footnote 5 Sometimes, when an organization could not accompany their post with a photo, they even decided to explain this to the public:

This project will support 375 needy children…They are located in many different towns and schools. Our workload is very heavy. Because we are now in the remote mountain areas, and mobile phones have no signal, we could not upload pictures in a timely manner.

Organizations also made quite a number of posts telling stories about exceptional volunteers (1.2 %) and educating the public about rural education or philanthropy in general (16.9 %). It should be noted that many of these posts are not directly related to the organization. For example, the volunteer featured in the story may not belong to the organization. However, these posts may still serve to promote the organizations; they can encourage other people to follow in the footsteps of the role models and participate in philanthropic work, or they can inform the general public of the severity of a certain social problem, and encourage people to take action.

Community: Engaging Netizens Using Their Language

Roughly a quarter of the Weibo posts we collected were community-building posts. In most of these posts, organizations greet newly acquainted peers, show their support for events organized by fellow NGOs, give thanks to donors and volunteers, solicit stakeholder opinions, and respond to questions raised by the stakeholders and the general public (Table 2). While the content of the these posts is usually serious, the tone of the posts can be very casual, full of popular Weibo slang, such as “darling (qin),” “little fellas (xiao huoban),” “awesome (geili),” “cute (meng meng da)” and “begging on my knees (gui qiu).” The use of emoji, such as “like (zan),” “heart (ai),” and “cry (ku),” is also common in the posts, reflecting the attitude and emotional states of the organizations (or rather the member of the organization that made the posts).

It should be noted that even in the informational posts, the use of slang and emoji is prevalent. This could be a reflection of the blogging habits of the individuals that made these posts. However, it could also be a strategic choice made by some organizations; when they are on Weibo and communicating with Netizens who prefer to use slang and emoji, the organizations need to perform accommodative code-switching so as to identify with their targets of communication and build credibility (Pullman 2013). Thus, although the contents of these posts are not interactive, their means of communication can still help the organizations to build relationships with others on Weibo.

Action: Promotion and Mobilization with the Newest Technologies

Only one eighth of the posts are action oriented, i.e., the post asks the followers to “do something” for the organization: make donations, become volunteers/employees, buy a product, participate in activities, cast a vote, or sign a petition (Table 2). While many of the posts call for action offline, when a particular action could be taken online, the post often contains a link or a Quick Response Code that directs the readers to the right address. Furthermore, with the newest e-commerce platforms such as Alipay and Wechat payment, many organizations not only give people the option to make donations online, but also encourage people to pay on their computer or by their smartphone. For instance:

#Making orders through Wechat is quite convenient! Highly recommend! Will we be overwhelmed by Wechat orders today?#

[hear][heart]@ WeiboID #Micro-philanthropy China# I donated for the project “rural children also have music dream!” through Micro Benefit China. The project needs 740.25 yuan in total. Now we are still 560 yuan short. Let’s help these lovely children together. Plant the seeds of hope! May all rural children have a happy childhood! @Micro Benefit China

Such posts seem to suggest that the Chinese NGOs are quite quick to adopt new technologies.

As shown in Table 2, mobilizing financial and human resources and promoting organizational activities are the most popular purposes of these action-oriented posts. We have collected only a few posts that are advocative in nature. No posts in the entire observation period called for a protest or rally. This is partially because of the nature of the NGOs in our sample. Another reason is that phrases such as “protest” or “rally” are considered sensitive words by the Weibo system. The system will automatically prohibit the publication of posts containing these words.

Posts Unrelated to Organizations’ Missions

Although the majority of the posts are closely related to the organizations’ work, there are a significant number of posts that are not related to rural education promotion or the field of philanthropy in general. For example, unrelated information constitutes 11.9 % of all posts collected, which is almost comparable to the entire Action category (12.5 %). Many of these posts are “chicken soup for the soul,” reflections on daily life, tips for cooking, jokes, domestic and international news, and so on. None has anything to do with the organizations’ mission. It is possible that, sometimes, an organization has made these posts (or has forwarded these posts) to attract the attention of followers, because these posts explore popular topics, and entertainment and gossip are a feature of Chinese social media. It is also possible that many Weibo accounts are operated by individuals, and they forget to maintain a boundary between their private lives and the organizations’ work. However, it is worth noting that among these posts, there are quite a number of posts containing news about the government, or content that openly praises the current government/government leaders, such as: “The achievement of Chairman Xi is positive energy for our whole nation!” and “To reform and to fight corruption, let’s see what Chairman Xi will do!” With the data at hand, it is hard to determine whether these posts are simply topics to attract followers, or gestures that the organizations made to impress their (potential) stakeholders from the government.

In the community-building category, aside from acknowledging events organized by other NGOs and sending holiday greetings, many organizations acknowledged memorial dates related to the anti-Japanese war. For example:

Respect the history, only sincere confession can win respect @ Japanese Embassy in Beijing.

#Today in history# September 9, 1945…the Japanese army formally surrendered. During the anti-Japanese war, 35 million Chinese soldiers and commoners died. China had made great sacrifice to finally come to this moment. We need to forever remember this day, and make ourselves stronger! @ China Youth Development Foundation. Never forget, and be strong! [Candlelight]@ CCTV News

#Memorial Day for the Defeat of the Japanese Army# [Wherever you are, please forward for those killed by the Japanese!] Sixty-nine years ago, August 15, 1945, Japan surrendered…Now the right extremists in Japan were trying to deny what they had done in China during the war…Every mountain and every river in China was wounded. This day could not be forgotten! Today, please forward this post to remember those who died, and to spread the truth! [Microphone]

To some extent, these posts added to anti-Japanese sentiment, and could hardly be counted as community building. However, since patriotism (or rather nationalism) is an important feature of the Chinese cyberspace (Wu 2006), spreading anti-Japanese information might also win support for the organization.

The original Information-Community-Action framework looks at how organizations uses social media functions. From the organizational legitimacy and resource mobilization perspective, it seems that organizations should be sharing mission-related information to attract followers, communicate with stakeholders about mission-related issues, and finally mobilize stakeholders to achieve organizational mission. However, our data suggests that organizations do not focus solely on their mission when it comes to communication on social media. Previously, Svensson et al. (2014) also observed that organizations were posting information unrelated to their programs on Twitter. However, they did not discuss reasons behind such a phenomenon. Our data is not able to tell us the reasons why the Chinese NGOs are doing so. We do not know whether they are making strategic choices to use these posts to attract followers, to show loyalty to the government, and to engage with the patriotic Netizens, or they are just being carried away by the culture of social media. These posts still fit into the Information-Community-Action framework, in that they are utilizing the information or dialog functions of Weibo. However, they seem to deviate from the legitimacy and accountability theories.

Discussion

This study is the first attempt to systematically understand Chinese NGOs’ behavior on domestic social media platforms. By observing the Weibo accounts of rural education NGOs, we show that Chinese NGOs’ social media usage pattern is both similar to and different from that of their Western counterparts. In our original sample, 36.53 % (179 in 490) of the organizations do not have a Weibo account, and 31.84 % (156 in 490) of organizations have a Weibo account, but made less than one post every 10 days. It is possible that some of the organizations have yet to adopt Web 2.0 technology. However, other factors need to be considered. A lot of the organizations in our sample have a presence on other interactive medium, such as Baidu Tieba, Renren, and QQ-zone. Some organizations maintain a blog or a BBS. If they have already established a good relationship with stakeholders using these platforms, they may not want to migrate to Weibo. Furthermore, in 2011, Wechat has emerged as a new social media platform. Because it is designed for the smartphone, it offers more innovative means of communication and interaction. With 355 million monthly active users, and availability in 200 countries, WeChat has become the most widely used social networking service in China (Lien and Cao 2014). Some organizations might have decided to use Wechat instead of Weibo. We have collected some posts in which the organizations asked their followers to follow them on Wechat. Hence, generally speaking, we believe the rate of social media adoption and the level of social medial involvement are likely to be high among Chinese NGOs.

When it comes to which functions of social media (information, community, or action) the NGOs use most, it seems that Chinese NGOs, like their Western counterparts, are mostly using social media as a one-way information dissemination channel, and that they rely on the information function more than their Western counterparts (63.7 % informational posts in our sample vs. 58.6 % in Lovejoy and Saxton’s sample vs. 47.8 % in Svensson et al.’s sample). To understand the Chinese NGOs’ heavy reliance on the information function, we need to first take into consideration the fact that in most of the previous studies on non-profits and social media the organizations being studied already had information-heavy websites (Lovejoy and Saxton 2012; Lovejoy et al. 2012; Svensson et al. 2014). In our sample however, a significant number of NGOs do not have a website (40 organizations, or 25.8 %), possibly, due to lack of resources. To them, a Weibo account is not an extension of an existing organizational website, but a cheap organizational website with interactive features. For other large and well-established NGOs, even though they have the resources to maintain a website, Weibo probably has a much larger audience than their individual websites. Furthermore, it is much easier for staff members to update Weibo than a traditional website on a frequent basis and it is much easier for a Netizen to follow news on Weibo than to refresh an organization’s webpage every few minutes (particularly, if he or she is following multiple organizations). Hence, some of these large organizations might have made a strategic choice to use Weibo as a replacement for traditional websites. Thus, rather than concluding that the Chinese NGOs have failed to utilize fully the interactive functions of Weibo, maybe we should say that the Chinese NGOs are trying to take full advantage of Weibo’s fast and convenient broadcasting function to make themselves heard.

What is more, from an organizational legitimacy perspective, information provision is the first step in negotiating organization legitimacy. The Chinese NGOs are facing some unique legitimacy problems, which might require them to spend more time and efforts providing information to their stakeholders. In our sample, at least one out of ten organizations has no legal identity. In other words, any activity these organizations engage in is considered illegal. To survive, they have to demonstrate to the general public, and maybe to government officials who are also surfing on Weibo, that they are trustworthy service providers rather than criminals. The large and well-established organizations, particularly those with a governmental background, face a different legitimacy crisis even though they have a legal identity. After a series of scandals in the charitable sector, the public’s trust in large charitable organizations has dropped significantly (China Charity Information Center 2011). Hence, these large organizations may attempt to use Weibo to rebuild their image.

As previous studies pointed out, even though social media platforms are less costly than traditional websites, organizations sometimes do not have the human resources to maintain an account. Thus they may open an account but failed to use its interactive features to the full. This might be true for some of the NGOs in our sample, as grassroots NGOs in China often do not have any full time staff, and their operation budget comes from contributions of their members (Zhou 2015). However, we would like to point out that there are some limitations in our data, which might depicted the Chinese NGOs as less active than they really are. For instance, Weibo is a hybrid of Facebook, Twitter and BBS. Although only the owner of an account can initiate posts, readers can comment by adding replies or forward the post with comments; and the owner of the account can further engage in discussion with the readers by replying to those comments under the original posts, or under the forwarded posts. As a result, even though the nature of the posts may not be dialogic, there might be a live discussion in the comment section or under forwarded posts. Due to technical limitations, although we have observed the above activities during data collection, we were not able to track them systematically. Moreover, we observed organizations using a large amount of slang and emoji in their posts. Whether this is conscious or unconscious, it may help the organization to identify with the general Weibo community. Thus, even though the primary purpose of some posts is not community building, they also serve a community-building role.

The failure to capture the more nuanced interactions between the NGOs and their followers also reflects the limitation of the Information-Community-Action framework. The framework is constructed based on the content of the posts, but not the tone of the posts; and it only counts the number of posts an organization makes, but not the responses (e.g., like, replies and retweets) of the followers. In previous studies, researchers have also noticed the limitation of such an approach. Lovejoy and Saxton (2012) for example, noticed the retweeting behavior, and called for studies on followers’ responses to the original posts. Other researchers found that when organizations use strategies to humanize themselves (e.g., informal language, or more expressive ways of communication), they decrease the distance between them and the audience, attract more followers (and thus creating a broader reach), and build a closer relationship with stakeholders (Denef et al. 2013; Lee et al. 2006). Based on our observations in China, as well as the findings of the previous studies, we suggest revising the Information-Community-Action framework. First, to reflect interactivities under the original or forwarded posts, each post should be weighed by the amount and type of responses generated. Second, a measurement of formality of the language, such as the F-score (Heylighen and Dewaele 1999, 2002) or the Coh-Metrix (Graesser et al. 2004) should be added to the framework. Then, we will not only be able to study what is being said, but also explore how things are being said.

Before making a conclusion, we would like to note that even though Weibo is heavily censored by the government, we only encountered three posts deleted by Weibo administrators. It is possible that with Weibo’s powerful filter system, the potentially offensive posts were prohibited from publishing in the first place. However, we believe that the major reason is the nature of our sample. As mentioned earlier, rural education is not a sensitive field, and it has received support from various sectors of society. As a result, the rural education NGOs are less likely to post material that might be deemed inappropriate and be removed by Weibo administrators. Other NGOs working in sensitive fields such as HIV, labor rights, human rights, homosexual issues, and religion, may behave differently on Weibo. Our findings cannot be generalized to these organizations. For other fields similar to rural education, such as poverty alleviation, medical assistance, and public health promotion, we believe that our findings are generalizable.

To conclude, in this study, we have shown that the Chinese NGOs, like their Western counterparts, are using social media to share information, build community, and mobilize resources, and that they use a large amount of slang and emoji to fit into the Weibo community. Despite Internet censorship and the unique government-NGO relationship in China, the Chinese NGOs remain active online. Using Weibo, the NGOs try to improve organizational legitimacy, as well as to attract further participation. Based on our findings as well as existing literature, we recommend possible revisions to the Information-Community-Action framework which can be applied to other social media platforms and organizations in other societies.