Foreign scholars generally classify emergencies as crisis events. They believe that a crisis can bring a series of unpredictable and unconventional events, which threaten social security [12] . Domestic scholars believe that emergencies are sudden and unpredictable events with an abnormal nature. The “People’s Republic of China Emergency Response Law” clearly pointed out that “an emergency refers to natural disasters, accidents, public health incidents and social security incidents that occur suddenly, cause or may cause serious harm to society, and need to take urgent measures to deal with.” The definition from “People’s Republic of China Emergency Response Law” is adopted in this study after comparison with other explanations.

Compared with traditional environment, in mobile Internet environment, emergencies are defined as sudden, destructive, and collective events with faster spread and wider scope. First, given that emergencies are in the interests of the masses, information spreading through mobile Internet arouses community and national attention and response instantly. Second, most of the contents refer to the government. Information on the inappropriate behaviors of the government or government staff most easily cause emergencies. Third, subjects and methods are diverse. Under the condition of mobile Internet, information spread not only relies on TV and newspaper channels; everyone can acquire and spread information through WeChat, micro-blog, small videos, and pictures nowadays.

Political participation is an activity undertaken by citizens to influence government decision making [13] . A citizen owns the right to express and show legal and political behaviors [14] ; where they can participate politically, express political appeal, supervise government’s actions, and influence political decision making [15] [16] [17] [18] . In the mobile Internet era, citizens express their political participation mainly through the Internet [19] [20] .

Internet political participation activities mainly include Internet voting, political mobilization, dialogue, and supervision as well as participation in group events [21] [22] . Internet political participation involves all political activities of Internet users, interest groups, and network elite [23] [24] . Internet political participation primarily aims to influence government’s decision making through a series of political activities [25] [26] .

Some scholars take university students as research object and divide Internet political participation into two dimensions: network political participation behavior and Internet political participation consciousness [27] . Some scholars also consider participation approach and believe that Internet users mainly use the Internet for the following objectives: browse and pay attention to government news and political events, discuss political issues, publish political views, conduct online voting, express demands on political departments and staff, and initiate political activities [28] . Other scholars divide Internet political participation methods into political information acquisition, political content communication, and political actions according to Polate’s research on the Internet’s effect on political activities [1] .

On the basis of previous research, Internet political participation in this study is classified into three dimensions: political information acquisition, political information interaction, and political behavior expression.

Political information acquisition is the basis of Internet political participation. People browse and search information on public events through intelligent mobile terminals, such as micro-blog, WeChat, and news channel browsing, and then access information resources for political participation.

Political information interaction refers to people communicating with others about public events or political news through mobile Internet platforms, forwarding public events news, or publishing their opinions on the platforms.

Political behavior expression refers to the manner in which the public participates in public events or political activities through mobile Internet platforms for voting, petition, or communicating with government departments and staff or supervising and reporting the behaviors of the government departments and staff.

Government information disclosure is an administrative behavior and system through which national administrative authorities or authorized organizations make the production or acquisition of information public to people within legal scope, according to specific procedures and methods and in a timely manner. In China, “National general public emergency contingency plans” clearly points out that right after an incident, the government should publish related brief information to the public immediately and conduct follow-up work on the event handling [29] . Therefore, timely information disclosure after emergencies is not only an obligation but also a responsibility of the government.

Two main explanations for government trust exist in the literature. One explanation regards it as psychological state, and the other one considers it as evaluation [30] . Citrin [31] asserted that government trust is public confidence on the government to provide public services with justice, honesty, equality, and efficiency [31] [32] [33] . Government trust also refers to the public’s evaluation and judgment on government behavior, that is, people judge whether the government deserves their trust by expected evaluation. Therefore, government trust implies that after emergencies, the public continues to believe that the government’s behavior and purpose are reliable and that the government still attaches importance to public interests. Thus, the public maintains their positive confidence on and expectation toward the government.

Government trust features three dimensions: ability, benevolence, and integrity [34] . 1) Ability pertains to the public’s evaluation on recognition and affirmation of the government’s ability to perform its duties, provide public services, assume public responsibility, solve public difficulties, and achieve public goals. 2) Benevolence refers to the public’s evaluation on recognition and affirmation whether the government carries out the work or makes a decision based on people’s interest, including whether the government behavior is consistent with the public interest, whether the government disclosures accurate and reliable information to the public, and whether the government is honest. 3) Integrity represents the public’s evaluation on recognition and affirmation on whether the government’s behavior is consistent with public morals and code of conduct, including whether relevant information is disclosed, whether the government work process is fair, and whether the government and its staff abide by their commitment or take responsibility.

Compared with traditional media, people are more willing to use the Internet to participate in political activities because the Internet offers technical advantages for information acquisition, dissemination, and interaction [35] . The Internet can also reduce the costs involved in political participation. Although people use Internet for chatting, entertainment, and reading apart from political participation, with the rapid expansion of Internet political influence, the government pays increasing attention to the construction of Internet political environment and government Internet platforms (e.g., the government’s official website, WeChat, and micro-blog) to gradually improve and strengthen political participation [36] . Thus, an increasing number of people communicate with government departments and participate in political discussions through the Internet. Moreover, the mobile network is rich in information resources which positively affect people’s online and offline political participation [37] [38] .

Political information is not only an important resource for public participation in politics but also an important dimension of political participation. Political information is closely related to government trust. Studies show that the public’s acquisition of emergency information through the Internet significantly affects their evaluation on the government’s capability of handling emergencies [39] . As one of the dimensions of government trust, information dissemination through the Internet significantly affects the evaluation of the government’s efficiency. The more positive information on the government’s work efficiency the people acquire from media, the higher they assess the ability dimension of government trust [40] . By contrast, the less political information the people acquires, the lower their trust on the government and staff [41] . The public generally believes that the government controls the preparation and management of social resources and the society. Therefore, when emergencies, such as food safety problems and environmental pollution, occur, the public generally attri- butes the causes to the government and staff’s lack of benevolence and integrity to provide reliable services and take full responsibility [17] [42] . Therefore, we put forward the following hypotheses:

H1a: Political information acquisition through the Internet after emergencies positively affects the ability dimension of government trust, that is, the more political information the people acquire through the Internet, the higher they trust the government’s ability.

H1b: Political information acquisition through the Internet after emergencies negatively affects the benevolence dimension of government trust, that is, the more political information the people acquire through the Internet, the lower they trust the government’s benevolence.

H1c: Political information acquisition through the Internet after emergencies negatively affects the integrity dimension of government trust, that is, the more political information the people acquire through the Internet, the lower they trust the government’s integrity.

Some scholars study the effect of media use on government trust, and research confirms that the longer and higher the frequency people use traditional media, such as TV and newspaper, the lower their evaluation of government trust. The media contains much negative information on government departments and officials. Thus, the more contact the public has with such contents, the lower their trust on the government. Negative information, especially rumors, spread fast and wide in the Internet environment after user sharing and interaction. Therefore, such information is likely to mislead the public’s perception of government behavior, resulting in the public dissatisfaction and doubt on the government [43] [44] .

Political information interaction is an expression of the public. In specific situations, people discuss and communicate about emergency development and the government’s handling of those situations through mobile Internet platforms, and this interaction directly influences evaluation of government’s ability. The public tends to expose mobile Internet platforms government’s negative news and complain about the government’s efficiency, government departments’ or staff’ unfairness, misconduct, or inability to provide satisfactory public service [45] [46] [47] . During these incidents, such as the “Shanghai Bund stampede” in 2015, “Harbin warehouse fire accident,” “explosion accident of Tianjin port,” and “Qing’an gun case,” Internet users’ sharing and interaction about these emergencies through WeChat, micro-blog, and other platforms influence the rapid spread of the incident. In addition, the local government dealt with those emergencies and public opinions slowly and inappropriately, and that official media literacy lagged behind; thus, the government’s ability to respond to emergencies is seriously questioned by the public on Internet platforms. Therefore, we present the following research hypothesis:

H2: Political information interaction through the Internet after emergencies negatively affects the ability dimension of government trust, that is, the more people interact about political information through the Internet, the lower their trust on the government’s ability.

Political behavior expression is the basic right of every citizen and a prerequisite for participation in political activities, such as political life, democratic governance, and management of public affairs. Political behavior expression is also the basis for ensuring correct operation of government authorities. The public owns the right to criticize and present opinions and suggestions on the behavior of the government departments and staff, and express their approval or opposition to the decisions made by the government [48] .

Normally, the public generally attributes their dissatisfaction on life quality and personal demands to the government. The reason lies in the hands of critical citizens. They usually maintain a skeptical and critical thought on those in authority and the regime. Spreading through the media such thought and disobedient attitude directly influence people’s thoughts and behaviors, thus the decline of people’s trust on the government [49] . The public’s suggestions, supervision, or reporting of the government through mobile Internet mainly aims to express negative evaluation on the government and staff’s working ability while looking forward to improved performance [50] [51] . That is to say, the more suggestions, supervisions, and reporting against the government that people expressed on mobile Internet, the lower is their trust to the government. In turn, many critical citizens actively respond on those thoughts once published and shared on mobile Internet, thus decreasing people’s trust on the government. Therefore, we develop the following hypothesis:

H3: Political behavior expression through the Internet after emergencies negatively affects the ability dimension of government trust, that is, the more political behavior the people express through the Internet, the lower their trust on the government’s ability.

Government information disclosure is an effective means to meet people’s information demands. Emergencies occur suddenly, develop rapidly, and are difficult to handle; thus, the government and the public are supposed to take specific measures to deal with emergencies [52] . Considering that emergencies threaten the lives and property of people as well as the stability of social security, people are eager to obtain relevant information on emergencies [53] . The government is not only responsible for controlling accurate information about emergencies but also obligated to publish relevant information.

Thus, we formulate the following hypotheses (Figure 1):

H4a: After emergencies, government information disclosure positively moderates the relationship between Internet political information acquisition and the ability dimension of government trust.

H4b: After emergencies, government information disclosure positively moderates the relationship between Internet political information acquisition and the benevolence dimension of government trust.

H4c: After emergencies, government information disclosure positively moderates the relationship between Internet political information acquisition and the integrity dimension of government trust.

H5: After emergencies, government information disclosure positively moderates the relationship between Internet political information interaction and the ability dimension of government trust.

H6: After emergencies, government information disclosure positively moderates the relationship between Internet political behavior expression and the ability dimension of government trust.

In this study, data were collected using questionnaire survey. The survey relates

to the different occupations and groups of all ages in different parts of China, including students, teachers, and other practitioners with access to mobile Internet. In total, 510 questionnaires were distributed, and the final number recovered was 480; the recovery ratio was 94.12%. In the formal questionnaire design, an additional item, which opposes the meaning of one of the items, was added for screening the efficiency and truthfulness of the questionnaire. After checking and analysis, ineffective questionnaires were omitted, and 423 effective ones were retained; thus, the ratio of valid questionnaires was 82.94%. Table 1 lists the profiles of objects.

This research mainly involves three latent variables: Internet political participation, government trust, and government information disclosure. Internet political participation was measured in three dimensions: Internet political information acquisition (three items), information interaction (three items), and behavior expression (four items) [54] [55] [56] . Government information disclosure was measured in terms of transparency (three items), timeliness (four items), and accessibility (three items) [57] - [62] . Government trust was measured from ability (four items), benevolence (three items), and integrity (three items). The

scale was based on Mayer’s [34] research and modified by considering Zaheer [63] , Aaron Martin [64] , and Xu Guihong [65] ’s studies.

Cronbach’s α was calculated to test reliability and for checking internal reliability and consistency of the scales used in this paper [66] . The results from SPSS proved that Cronbach’s α of Internet political participation (0.832), Internet information disclosure (0.915), and government trust (0.920) are all greater than 0.70; thus, the reliability and validity of the scales are acceptable and reliable. Details are shown in Table 2.

Factor analysis was conducted to test the validity of scales and ensure that questionnaire contents and structure are effective [67] [68] . Kaiser [69] thought that KMO can be calculated to judge if the items of different scales are appropriate for factor analysis, that is, when the value of KMO is less than 0.5, the items are not suitable for factor analysis [69] . By contrast, if the value is greater than 0.80, then the items are suitable for factor analysis. The KMO value of Internet political participation (0.808), government information disclosure (0.874), and government trust (0.899) are all greater than 0.80. Furthermore, in view of Bartlett’s test, the value of p of all the variables is less than 0.05; thus, common factors among the variables and the variables are suitable for factor analysis. Therefore, the reliability and validity of all the scales in this study are acceptable and satisfactory. Details are given in Table 3.

Table 4 indicates that the correlation coefficient between Internet political information acquisition and the ability dimension of government trust is 0.209**. This value signifies that Internet political information acquisition is positively related to the ability dimension of government trust. In addition, Internet political information acquisition is positively related to the benevolence dimension

Note: *p < 0.05, **p < 0.01.

(0.095*), but negatively related to the integrity dimension (−0.105*). By contrast, Internet political information interaction is not significantly related to government trust, and Internet political behavior expression is negatively related to the ability dimension (−0.142**). Moreover, government information disclosure is significantly related to both Internet political participation and government trust. The correlation analysis in Table 4 can partially prove the hypothesesH1a, H1c and H3. Thus, multiple linear regression was conducted to further test the hypotheses.

Table 5 presents the results of multiple linear regression. In the regression, control variables and independent variables were added in the models gradually to test the relationship with government trust. To avoid multicollinearity of the function between Internet political participation and government information disclosure, the two variables were decentralized.

In the regression results of the ability dimension, F = 6.38*** in model 1, indicating that control variables, including age (β = 0.258***), political status (β = 0.325*), and occupation (β = 0.114***), exert significantly positive effect on the ability dimension of government trust. In model 2, Internet political information acquisition is positively related to the ability dimension (0.251***), whereas Internet political behavior expression (−0.149**) is negatively related to the ability dimension, and Internet political information interaction exerts no apparent effect on this dimension (0.037). Therefore, H1a and H3 are supported, but H2 is

Note: 1. (X − X ¯ 1 )(Z − Z ¯ ), (X − X ¯ 2 )(Z − Z ¯ ),(X − X ¯ 3 )(Z − Z ¯ ) are functions of the three dimensions of internet political participation and government information disclosure. 2. *p < 0.05, **p < 0.01, ***p < 0.001.

not. In model 4, the regression result of the function of Internet political information acquisition and government information disclosure on the ability dimension of government trust is significant (0.111**), implying that the effect of Internet political information acquisition on this dimension is positively moderated by government information disclosure, that is, the effect is stronger when more relevant information is published by the government. Thus, H4a is supported. Similarly, the function of Internet political behavior expression and government information disclosure is positively related to the ability dimension of government trust (0.122***). On the contrary, the moderating effect of government information disclosure is insignificant on the relationship between Internet political information interaction and the ability dimension of government trust (−0.083); thus, H6 is confirmed, but H5 is not.

For the regression results of the benevolence dimension, some control variables, including age (0.210***) and occupation (0.160***), exert significant positive effects on this dimension, as observed in model 1. Only Internet poli- tical information acquisition significantly affects the benevolence dimension (0.134**), which indicates an opposite result against H1b. Internet political information interaction and Internet political behavior expression are not significant to the benevolence dimension of government trust. In addition, given that the regression result of the function of Internet political information acquisition and government information disclosure is significant (0.122**), government information disclosure positively moderates the relationship between Internet political information acquisition and the benevolence dimension. Thus, H4b is confirmed.

For the regression results on the third dimension of government trust, age (0.315***), political status (0.352*), and occupation (0.253***) exert positive effects on integrity in model 1. Internet political information acquisition is negatively related to integrity trust (−0.111**). Thus, H1c is confirmed. In addition, the positive moderating effect of government information disclosure on the relationship between Internet political information acquisition and the integrity dimension of government trust is also evident in model 4 (0.079*). Thus, H4c is supported.

The following are the several contributions of this study.

First, we studied the effect of unexpected events on government trust using mobile Internet environment as basis. This research not only reflects the era of mobile Internet, but also reveals real challenges brought by mobile network technology.

Second, we analyzed and classified the concept and dimensions of Internet political participation in mobile Internet environment and tested how Internet political participation affect government trust. Results confirmed that two dimensions significantly influence government trust.

Third, in contrast with many scholars’ view on government information disclosure as an independent variable, we constructed a relationship model between Internet political participation and government trust, regarded government information disclosure as moderator, and tested the model.

Using mobile Internet, emergencies, political participation, Internet political participation, government information disclosure, trust, government trust, and other related theories as basis, we proposed the relationship model of Internet political participation, government trust, and government information disclosure after emergencies. Empirical analysis proved that most of the hypotheses were verified. The following three points should be further improved.

First, the objects of study should include people in civil service, military, and judicial system, that is, the so-called institutional personnel. The next step is to study the effect of “institutional” and “non-institutional” staff’s trust on the government. Different empirical research objects may exert different effects on the results.

Second, people in 11 provinces and cities nationwide were involved in the research to make the sample scientific, and most of the questionnaires were completed by our friends and classmates. They might not understand the questions clearly. Hence, data quality may not be that satisfactory, and the conclusions may be limited.

Third, the index of the variables in this study is mainly based on existing research and modified through literature review and theoretical revision without objective data support. Thus, further theoretical and empirical research is needed to improve the variable scales.