Introduction

Trust is a fragile asset, easy to tear down yet difficult to build up. Researchers have long recognized the value of trust in legal authorities in maintaining the legitimacy and effectiveness of a state’s formal social control apparatus (Decker 1981; Sunshine and Tyler 2003). Legal institutions, such as the police and courts, have widespread authority to use coercive force and adjudicate punishment, and must subject their power to public scrutiny in a democratic society. Citizens who trust legal authorities are more likely to comply with the law and accept dispositions of criminal justice practitioners voluntarily (Tyler 1990).

The purpose of this study is to compare public trust in legal authorities in two Chinese societies: China and Taiwan. During the past two decades, China has emerged as a global economic and political power, and arguably one of the more stable authoritarian regimes in the world. The rapid social and economic transformations, however, have resulted in a gradual erosion of the state’s coercive power and a legitimacy crisis of its criminal justice system (Tanner 2004; Wong 2004). Taiwan, meanwhile, has also gone through dramatic social and political changes since the 1980s and has become one of Asia’s most developed democratic societies. In recent years, however, facing waves of economic recessions, coupled with corruption scandals in high places, the authorities suffer attenuated legitimacy and public support.

Despite a growing body of research on public attitudes toward legal authorities in Chinese societies (e.g., Cao and Dai 2006; Lai et al. 2010; Landry 2008, 2011; Michelson and Read 2011; Wu and Sun 2009, 2010), a critical area of inquiry regarding the effects of the media is missing. This omission is problematic as media is an important source from which most people derive their knowledge about legal authorities. Indeed, research in the West has shown that the media play a significant role in public perceptions of government in general (Miller et al. 1979) and legal institutions in particular (e.g., Callanan and Rosenberger 2011; Dowler 2002; Indermaur and Hough 2002; Johnson and Bartels 2010; Roberts and Hough 2002; Surette 2007; Weitzer and Tuch 2006; Wu et al. 2010). This study aims to fill that void.

Relying on nationwide survey data, this study examines: (1) to what extent do Chinese and Taiwanese citizens trust their police and courts; and (2) how does media consumption influence Chinese and Taiwanese trust in police and courts, after controlling for a range of individual demographic, attitudinal, experiential, and locality variables? Separating the police and the court in analysis is helpful as the two institutions vary significantly in their degree of visibility, level of contact with the public, and scope and type of media coverage, possibly leading to differential public sentiments and perceptions.

Chinese Evaluation of Legal Institutions

The relationships between legal authorities and the people in China and Taiwan have undergone noticeable changes over the past several decades. In mainland China, although traditional values emphasize respect for authority and communist policies highlight serving the people, the relationship between legal authorities and the public may be strained. Compared to police in Taiwan and Western countries, Chinese police are given much greater authority and power (Sun and Wu 2010). The unequal distribution of power renders the police as the superiors and citizens as the subordinates (Jiao 1995). In recent years, police misconduct has emerged as a serious social problem (Wong 2004). Between 1993 and 1997, more than 40,000 police officers were involved in over 33,000 criminal cases or disciplinary violations, with an average of 8,000 officers or 6 % of total officers punished annually (Wang 2007). The effects of governmental efforts to control police misconduct are likely to be limited though, due to, among other factors, the absence of respect for the law in police culture, the over-reliance on monetary incentives to perform police work, and the reluctance of the government to expose large-scale police corruption and problems (Dutton 2005; Sun and Wu 2010).

Police scholars have noted that Chinese citizens today are more willing to challenge legal authorities than before (Tanner 2001; Wong 2002). As economic and social inequality worsens in the past two decades, conflicts within the Chinese society escalate; public demonstrations and protests grow in both number and size, and negative contacts between the police and citizens increase (Mainland Affairs Council 2005; Tanner 2004). Occupational risks for the police have been rising (Sun and Wu 2010) and the increased danger of their working environment may change officers’ mentality when interacting with citizens. A recent study found that Chinese police cadets were more likely than their counterparts in the United States to distrust citizens (Sun et al. 2010).

Despite these negative developments, the majority of Chinese citizens rated police favorably. An early study showed that more than 80 % of the respondents reported that Chinese police did a good job in controlling crime (Zhu et al. 1995). A recent study found that 72 % of the respondents trusted police a lot or to a degree (Wu and Sun 2009), and another study reported that 56 % of the surveyed college students were “very satisfied” or “satisfied” with police (Wu and Sun 2010). Comparative research on China and Taiwan showed that Chinese have higher levels of trust and confidence in their police than do their Taiwanese counterparts (Lai et al. 2010; Wu et al. 2010).

Compared to the police, courts are much less visible and more remote from people’s lives. While “When in need, find the police” has been a popular slogan among the Chinese, the old proverb “It is better to be vexed to death than to bring a lawsuit” still goes a long way (Di and Wu 2009). As the vast majority of Chinese citizens do not have direct contacts with the courts, stories and experience shared among families, friends, and acquaintances and through the media may become especially important sources of information.

In the past two decades, although the Chinese courts have made a great deal of progress in improving judges’ training and professionalism and ensuring procedural regularity, there remain a number of problems that can significantly reduce public trust. The Chinese judicial system is described as “more a system of discretion supplemented by law than a system of law supplemented by discretion” (Woo 1999, p. 615). Although judges have less authority compared to their Western counterparts, they have ample opportunities to exercise self-interested discretion with their broad responsibility in collecting evidence and conducting investigation. Personal appeals are used frequently to obtain favorable judicial outcomes, and those who have power, connections, and money get ahead (Woo 1999). Further, Chinese courts suffer more external interference (Belkin 2000). Individual judges do not have legally recognized independence (Li 2003) and are often influenced by directives from the Party and government officials and the needs of powerful and rich individuals (Li 2003). As local courts depend on local governments for financial support and the appointment and promotion of judges, local judicial protectionism is a serious problem, threatening the integrity of judges (Lubman 2003). All these problems may abate the authority of the courts and undermine public trust.

Like their evaluations of the police, Chinese citizens’ general perceptions of the courts are highly favorable. Using response categories ranging from −2.5 to 2.5, a study found that the average response value for the question about confidence in the courts was 1.13, much higher than the midpoint of 0.0 (Chen and Shi 2001). Similar patterns of positive ratings of the courts were reported in a study of Chinese urban and rural residents’ perceptions of the courts (Michelson and Read 2011) and a comparative study of Chinese and Taiwanese confidence in courts (Lai et al. 2010).

In Taiwan, fundamental changes in the criminal justice arena have occurred in the past three decades. During the martial-law years (1949–1987), the police were controlled tightly by the Nationalist Party (Kuomintang or KMT) and the central government to repress political dissents and promote regime legitimacy. After the lift of the martial law in 1987, demonstrations and protests were permitted, strikes and collective bargaining were legitimized, and violent confrontations between citizens and police decreased greatly. The police face new challenges with the democratic transition, however. As the principles of the rule of law and protection of human rights are underscored by democratic policing, the police are now expected to have higher standards of performance and subject to stronger external control and scrutiny by local officials, citizens, and the media.

In the martial-law era, the Taiwanese courts served as a powerful tool for the KMT to maintain political control by oppressing political dissents and protecting corrupt KMT members (Wang 2002). The legalization of new political parties and the establishment of a two dominant-party system, however, have made the courts more independent and democratized. The political power is now subordinate to the law, and human rights are proclaimed to be the ultimate value of the Taiwanese society (Chen and Lo 2010). Judges are more professionally trained (Wang 2010). The inquisitorial model is transformed into a reformed adversarial model (Wang 2002). More attorneys are recruited to provide legal services to the people (Chen and Lo 2010).

The Taiwanese judiciary, however, is not without problems. The issue of equal justice to all remains a concern. While the Taiwanese people enjoy high standards of legal procedure at higher-level courts, the ideals of the rule of law are not similarly carried out in the lower courts. Litigants who have more resources of knowledge, wealth, and influence are more capable of using the courts than the “have-nots” (Chen 2009). Corruption is also a problem. The recent case involving three senior High Court judges taking bribes is considered one of the worst corruption scandals in Taiwan’s judicial history, and the head of the judiciary was forced to resign as a consequence (Sui 2010). The need to establish a better qualification system to select judges, an effective mechanism to monitor judge performance, and an open environment for competent judges to deliver independent judgments remains to be addressed in Taiwan. Evidence showed that Taiwanese citizens were much more critical of their courts compared to Chinese citizens (Lai et al. 2010).

Media in China and Taiwan

China and Taiwan are good comparison sites with important contrasting features to investigate the effects of the media on public perceptions. In China, during the pre-reform era, the media served as the mouthpiece of the regime, manipulating public opinions and facilitating policy implementation (Liu 1971). The post-Mao reforms have put the media through gradual, and sometimes, profound changes, including liberalization and diversification (Chen and Shi 2001; Polumbaum 1990). Many formerly forbidden areas, such as negative societal news, human interest news, and critical legal reports, are frequently broadcast (Zhao 1998). New sources of information, such as the Internet, are growing rapidly. In 2008, Chinese net users have exploded to over 253 million, surpassing the number of users in the United States (Esarey and Qiang 2008).

The Chinese government, nonetheless, keeps control of the media and continues to use propaganda to promote support and discourage criticism (Chen and Shi 2001; Yang and Tang 2010). News reported by state-owned television stations and newspapers are replete with accounts of exemplary police officers and judges in an effort to promote positive images of these authorities. The government also exercises tight control over the Internet and employs a wide range of methods to promote self-censorship among journalists and the whole media community (Hassid 2008). The actual effects of such mobilization efforts on public attitudes, however, is unclear. Zhu (1990) argued that the effects of government propaganda vary by areas: it is most effective in repressing dissent, somewhat effective in promoting government policies, and least effective in making people trust the authorities. Chen and Shi (2001) actually found that news media in China has negative effects on public attitudes toward political institutions, making people more distrustful of the government.

The media in martial-law Taiwan was subject to similar political interference as in China, delivering government propaganda information and politically correct entertainment (Huang 2009). After martial law was lifted, the media underwent fundamental changes in content, structure, and function. The majority of news agencies in Taiwan today are no longer state-owned. To entice audience, different agencies, private- and state-owned alike, actively seek for sensational news on the government and public figures (Rawnsley 2004). Negative, sometimes exaggerated, reports of official misconduct may damage the authorities’ image severely and lower citizens’ trust. One recent study found that the time spent on newspapers was related negatively to political trust in Taiwan (Chen and Lo 2010). More research is needed to fully discover the effects of the media on public opinion on legal authorities in both China and Taiwan.

Media and Public Trust

Existent evidence, based mainly on data collected from the West, suggests that the connection between media consumption and trust is complex, and needs to take a number of factors, including the type and content of the media, frequency of media exposure, and attitudes toward the media, into consideration.

Different types of media may have uneven influences on people’s trust in legal authorities. For example, reading newspapers was found related to higher trust in government institutions and watching television news was found associated with lower political trust and greater political cynicism (Moy and Pfau 2000). In a recent study of public support for American courts, Johnson and Bartels (2010) found that sensationalist media exposure (to political radio and cable news) depresses public support for American courts, whereas sober media exposure (through newspapers and network news) has no such effects.

Except for newspapers and televisions, we know little about how other media types may affect public trust in legal authorities. New types of media, the Internet in particular, have greatly changed the landscape of the industry, with an increasing number of people relying on the Internet for news. A recent study revealed no significant impact of exposure to online campaign news on political trust in the United States (Avery 2009), but more research is needed to understand the effects of the use of the Internet on trust in legal institutions.

Besides the type of the media, the content of the media matters. Robinson’s (1976) malaise-impact theory posits that perceived credibility of television networks and negative emphasis of television news coverage lead to political cynicism and mistrust in democratic societies. Heavy viewers of network news and crime shows are found more likely to perceive police misconduct as a frequent occurrence (Dowler and Zawilski 2007). News reports of incidents of police misconduct are also found to reduce public satisfaction with the police (Kaminski and Jefferis 1998; Weitzer 2002). Frequent exposure to such negative news reports has a particularly strong impact on citizens’ satisfaction with the police and their beliefs in the frequency of police misconduct (Weitzer and Tuch 2006).

Besides the type and the content, the frequency of exposure to news may influence public trust in legal institutions. Norris’s (2000a, b) virtuous circle theory postulates that frequent media usage can lead to higher levels of trust in government and civic engagement, contrary to what the malaise-impact theory proposes (Robinson 1976). He argues that people who are more politically interested, engaged, and trusting will watch more political news, learn more about the government and politics, and accordingly, cultivate greater trust and engagement, producing a virtuous circle that enhances democracy. Empirically, Norris (2000b) found that confidence in political parties and politicians went up in Great Britain during the 1997 election campaign when public attention to political news was also at its height. It is unclear, though, if consuming political news has the same trust-promoting effects in China. It is also unknown if exposure to different areas of news, such as domestic and foreign news, has a similar trust-enhancing effect.

Lastly, people’s trust in media can possibly influence their trust in legal authorities in two ways: (1) both the media and the police and courts are visible public institutions that may evoke similar sentiments from the public; and (2) the effects of the media may be contingent upon people’s initial trust in the media (Miller and Krosnick 2000). That is, for those who are initially less trustful of the media, media exposure may have a minimal or null effect on their political trust, but for people who are initially trustful, news media may increase their political trust significantly (Chen and Shi 2001).

Data and Methods

Data Source

This study used data collected by the second wave of the Asia Barometer Survey (ABS) (2006–2008) to examine the effects of media consumption on public trust in legal authorities in China and Taiwan. Conducted annually since 2003 in 17 countries in East and South Asia, the ABS is the largest and most systematic survey of public opinion on political values, democracy, and governance in the region. Currently based at the National Taiwan University, the project has a national research team in each country to administer face-to-face survey interviews with citizens. A common research framework, survey module, and methodology are employed in all countries to ensure that cross-national data are reliable and comparable.

The China data were collected in 2008, using stratified multistage cluster (area) sampling techniques. The respondents were adults aged 18 and above, including citizens from all 22 provinces, four autonomous regions, and four municipalities, but excluding the Tibet Autonomous Region due to limited resources.Footnote 1 The sampling procedure started with a selection of 67 primary sampling units (PSUs) based on their location, geographic characteristics and region size. These PSUs included counties in rural areas, cities in urban areas, and counties in municipal areas. From each of the PSUs, second sampling units (SSUs) were selected, including township, area and street. After that, administrative villages, urban resident committees, and local community committees were selected from each of the SSUs. Finally, households were selected from these villages and committee areas. The number of households in each county and city was determined by the probability proportional to size (PPS) sample designs.

The Taiwan data were collected in 2006, following a similar procedure of stratified multistage cluster sampling. First, a total of 328 sampling units, together with Taipei City and Kaoshiung City, were clustered into ten groups. The number of samples in each group was determined by the PPS measures. Second, systematic sampling was performed separately within each group to obtain primary sampling units (PSUs). The same procedure was employed to obtain the secondary sampling units (SSUs). Finally, respondents were selected randomly from each SSU. Voting-age adults (21 and above) were eligible to participate in the study.

Except for a few additional questions that the China survey included, the two surveys have identical items. Pre-tests were conducted to correct unclear wording and vague sentences. Rigorous quality control efforts were made before, during, and after the distribution of the survey. Interviewers were intensively trained, assessed, supervised, and spot-checked.Footnote 2 In China, Mandarin was the official interview language. In Taiwan, both Mandarin and Taiwanese were provided as options. Samples were weighted to better represent the population in the structure of age, education, urban/rural distribution (China), and geographic distribution (Taiwan).

Measures

To capture trust in legal authorities, respondents are asked how much trust they have in the police and courts (1 = none at all, 2 = not very much trust, 3 = quite a lot of trust, 4 = a great deal of trust). The two variables are recoded into dummy variables (0 = no trust or not much trust, 1 = quite a lot or a great deal of trust) for binary logistic regression as the assumption of parallel lines required for ordinal logistic regression is not met.

The independent variables measure different aspects of media consumption. The first two variables indicate the use of television and newspaper as the main source of information. Respondents are asked whether televisions are their main source of information (0 = no, 1 = yes) and whether newspapers are their main source of information (0 = no, 1 = yes). The third variable indicates the frequency of the use of the Internet (1 = not aware of Internet or never, 2 = hardly ever, 3 = several times a year, 4 = at least once a month, 5 = at least once a week, 6 = almost daily).

Two variables measure the frequency of people’s consumption of domestic and international news. Respondents are asked how often they follow news about politics and government (1 = practically never, 2 = not even once a week, 3 = once or twice a week, 4 = several times a week, 5 = everyday) and how closely they follow major events in foreign countries/the world (1 = not at all, 2 = very little, 3 = not too closely, 4 = somewhat closely, 5 = very closely). The last media-related variable, trust in media, is indicated by an additive scale of two items asking respondents how much trust they have in newspapers and televisions respectively (1 = none at all, 2 = not very much trust, 3 = quite a lot of trust, 4 = a great deal of trust).

Four groups of theoretically relevant variables are controlled in this study, including demographic attributes, social and political trust, experience with and perception of crime, and residential areas and country. Demographic variables include age (measured in years), gender (0 = male, 1 = female), marital status (0 = not married, 1 = married including living-in as married), education (10 categories ranging from 1 = no formal education to 10 = post-graduate degree), and perceived socioeconomic status (SES), indicated by a ten-point scale on which the respondents self-rank their SES in the society.

Social trust is gauged by an additive scale of three items indicating how much trust the respondents have in their relatives, neighbors, and other people they interact with (1 = not at all, 2 = not very much trust, 3 = quite a lot of trust, 4 = a great deal of trust). Political trust is indicated by respondents reporting if they agree that they can generally trust the people who run their government to do what is right (1 = strongly disagree, 2 = somewhat disagree, 3 = somewhat agree, 4 = strongly agree). This variable, indicating people’s trust in incumbent governmental officials, is conceptually different from trust in legal institutions, such as police and courts (Chen and Shi 2001). Another variable, corruption, is included to indirectly measure political trust, asking respondents whether they themselves or anyone they know have personally witnessed an act of corruption or bribe-taking by a politician or government official in the past year (0 = no, 1 = yes).

Two variables signal people’s sense of safety and experience of victimization. Sense of safety is a factor of two items (factor loadings =  0.78). Respondents are asked “Generally speaking, how safe is living in this city/town/village?” (1 = very unsafe, 2 = unsafe, 3 = safe, 4 = very safe) and “Compared to the situation in this city/town/village a few years ago, do you feel more safe, less safe or the same as before?” (1 = less safe, 2 = same as before, 3 = more safe). To measure victimization, the respondents are asked whether they have been a victim of one or more of four types of crime: car, motorcycle, or bicycle theft, pick-pocketing/robbery of personal property, break-in at home, and physical violence (0 = not victimized, 1 = victimized).

Finally, regarding residential areas, three dummy variables are created representing respectively capital/megacities, small towns, and village/countryside, with capital/megacities serving as the reference group. Country is indicated by a dummy variable (0 = Taiwan, 1 = China). Table 1 displays the descriptive statistics for all the variables used in this study. Potential multicollinearity problems were checked by examining the variance inflation factors (VIFs) and all the VIF values are below the conservative cutoff value of 2.5 for binary logistic regression (Allison 1999). Two-variable correlation matrix analysis among all independent variables also confirms that multicollinearity is not a concern.

Table 1 Descriptive statistics for variables in regression analyses on trust in legal authorities
Full size table

Results

Table 1 shows the mean scores of the trust in legal institutions variables and Table 2 reports the percentage distributions of these variables. Two general patterns arise. First, Chinese respondents reported higher levels of trust in police and courts than did Taiwanese respondents. The mean scores of trust in police and courts were 3.07 and 3.11 for Chinese respondents, compared to 2.45 and 2.27 for Taiwanese respondents. A much higher percentage of Chinese respondents (36.4 %) reported a great deal of trust in police than Taiwanese respondents (6.3 %). Only 27.8 % of Chinese respondents reported either not very much or none at all trust in police, while over half of Taiwanese respondents reported so. Similarly, 37.9 % and 38.7 % of Chinese respondents stated that they have either a great deal of trust or quite a lot of trust on the courts, compared to only 3.8 % and 30.7 % of Taiwanese respondents who answered so. Second, comparing trust in the two legal institutions, Chinese respondents expressed more trust in courts than police, whereas Taiwanese respondents displayed more trust in police than courts.

Table 2 Percentage distributions for trust in legal authorities
Full size table

Table 3 presents the results of binary logistic regression on trust in legal authorities. Both combined and separate analyses were performed to reveal the potential similarities and variations between the Chinese and Taiwanese respondents.

Table 3 Binary logistic regression results on trust in legal authorities
Full size table

Analyzing Chinese and Taiwanese respondents as a combined sample, one media variable, trust in media, exerted significant effects on trust in police. People who trusted the media were more likely to trust the police. Several control variables were significant predictors of trust in police. People who had higher levels of perceived social status and who resided in village/country side were more trustful of police compared to their counterparts who had lower self-perceived SES and who lived in large cities. Regarding attitudinal variables, both social and political trust were associated with higher odds of police trust and recent witnessing of corruption or bribe-taking by a politician or government official was related to lower odds of trust in police. In terms of crime-related variables, sense of safety had a significant positive effect on Chinese trust in police. After holding a variety of independent and control variables constant, the country variable remained significant—the odds of Chinese respondents reporting trust in police were over 50 % higher than those of the Taiwanese.

The analysis of separate models revealed that among Chinese respondents, those who followed major events in foreign countries/the world more closely had lower odds of trust in police. Exposure to foreign news, however, did not affect Taiwanese trust in police. Different from the Chinese respondents, Taiwanese respondents’ trust was not predicted by self-perceived SES, residence of village/countryside, or sense of safety. For both groups, trust in media, social trust, political trust, and corruption were significantly linked to trust in police.

With respect to trust in courts, the combined sample analysis indicated that ten variables had significant effects. Exposure to foreign news was negatively related to and trust in media was positively related to trust in courts. People who were married or who had higher levels of education had lower odds of trust in courts, while people who had higher SES had higher odds of trust. Witnessing corruption was associated with lower odds of trust in courts, whereas social trust, political trust, and sense of safety were linked to higher odds of trust. After holding all other independent and control variables constant, Chinese respondents showed substantially greater odds of trust in courts than did the Taiwanese. In fact, the odds of Chinese having trust in courts were almost five times the odds of Taiwanese, bypassing the discrepancy in trust in police.

Comparing the separate model analysis of the Chinese and Taiwanese respondents, five variables have differential effects between the two groups. Chinese respondents who closely followed major events in foreign countries/the world or who were married were less likely to trust the courts, yet exposure to foreign news or marriage did not influence Taiwanese trust. Similarly, sense of safety had a significant trust-boosting effect among the Chinese, but not the Taiwanese. Meanwhile, education and recent experience with corruption or bribe-taking were associated with distrust in courts for Taiwanese people, but not for Chinese. The effects of trust in media, subjective SES, social trust, and political trust were similar for the two groups of respondents.

Summary and Discussion

This study compares trust in police and courts in China and Taiwan, with an emphasis on the effects of media consumption. A few main findings deserve discussion. To start with, Chinese citizens have substantially higher levels of trust in legal authorities compared to Taiwanese citizens. This finding is consistent with the general patterns reported by two recent studies (Lai et al. 2010; Wu et al. 2010). One possible explanation ties to the critical citizen thesis, positing that democratic countries have seen a rise of critical citizens who are less trustful and deferential to authorities and more ready to challenge them (Norris 1999; Pharr and Putnam 2000). Living in a democratic polity, Taiwanese people are more likely to hold a skeptical view toward their legal authorities compared to Chinese who are living in an authoritarian society.

A second possible explanation for this attitudinal discrepancy is that Chinese respondents have higher levels of trust in media and trust in people who run the government (as indicated by the mean scores reported in Table 1)—two factors that show significant enhancing effects on trust in legal authorities in both China and Taiwan in this study.

There may be a third explanation, i.e., Chinese respondents do not express their true opinions due to heightened political fear or intimidation (Lai et al. 2010; Wu and Sun 2009). This study uses three methods to preliminarily test this possibility (similar approaches see Shi 2001 and Yang and Tang 2010). When people are unwilling to express their true opinions, they most likely would choose the “do not know” option, refuse to answer, or give false answer (Shi 2001). The first method thus examines the missing data where the respondents answer “can’t choose” or “decline to answer.” Results show that there is a higher proportion of missing data among the Chinese (6.7 %) than Taiwanese (4.7 %) respondents in the trust in police measure, but a higher proportion among the Taiwanese (11.4 %) than Chinese (9.0 %) respondents regarding the trust in courts measure. The second method gauges the magnitude of self-reported political fear and its correlation to trust in legal authorities. Respectively 89 % and 75 % of the Chinese and Taiwanese respondents report that they agree “People are free to speak what they think without fear.” If arbitrarily categorizing cases with missing data as disapproval of the statement, the attitudinal discrepancy is narrowed with Chinese (77 %) remaining slightly more positive about freedom of speech than Taiwanese (73 %). The correlations between political fear and trust in legal authorities are similar between Chinese and Taiwanese.Footnote 3 The last method relies on the assumption that if respondents fear about political prosecution, a positive correlation between reported trust and the presence of a third party in the interview site should be expected. Results show that there are no correlations between the presence of a third party, or the presence of a third non-family member party, and trust in legal authorities, in China or Taiwan. These preliminary tests seem to refute the heightened political fear thesis.

Comparing trust in the two legal authorities, Chinese citizens trust courts more than police, but Taiwanese citizens trust police more than courts. As the Chinese police are endowed with greater power and authority than the police in Western democratic countries and Taiwan, these findings may reflect the dissatisfaction of Chinese people with their police’s excessive power. In addition, police have a much more visible presence in community compared to courts, with the majority of police work operating on the street level and officers having the widest contacts with the public. As a result, police officers may receive much closer public scrutiny than courts, leading to lower public trust. Furthermore, it is possible that the media have reported more negative news on police misconduct than on courts, contributing to more negative views of the police in China.

The opposite pattern shown by Taiwanese respondents is similar to what is found in the United States. The Gallup Poll from 2004 to 2010, for example, consistently indicated that American citizens were more likely to have a great deal or quite a lot confidence in the police than in the Supreme Court or the criminal justice system overall (Sourcebook of Criminal Justice Statistics 2010). It seems that in democratic societies, as the police’s mission is to protect and serve the citizenry, as opposed to “to intimidate and suppress opposition and to protect the regime” in authoritarian societies (Wiatrowski and Goldstone 2010, p. 79), the police tend to receive more public support than courts, an institution that keeps a certain distance from the public and is in charge of imposing forms of sanctions that arguably have a more severe impact on citizens than the police.

Somewhat unexpected, media consumption only has a limited effect on public trust in legal authorities in both societies. Results show that consumption of televisions, newspapers, and the Internet does not influence either Chinese or Taiwanese trust in police or courts. The nil effect of different media types suggests that the Chinese government’s propaganda efforts, aiming at nurturing supportive sentiments toward the authorities, may have little actual impact on public opinion on the authorities.

This result, however, is not completely unexpected. Despite governmental attempts to manipulate public attitudes through monopolistic control of the media and strict news censorship, the media has become more commercialized and decentralized in present China, and is no longer just a propaganda tool for the Party or the government. To survive, the media has to cater to the wants and needs of an increasingly diversified and sophisticated audience (Lynch 1999). Indeed, despite government control, investigative journalism has been on a rise, with growing numbers of television programs, newspaper reports, and particularly Internet articles and discussions criticizing social and legal problems, including police and judiciary misconduct (Liebman 2005). The mixed messages delivered through these media channels may undermine the trustworthiness of legal authorities.

Interestingly, this study finds that the frequency of exposure to foreign news has a differential effect on Chinese and Taiwanese trust in legal authorities. While exposure to foreign news does not influence Taiwanese trust, Chinese citizens who pay more attention to major events in foreign countries are less likely to trust police or courts. This study proposes a “contrast” hypothesis to explain this phenomenon. It is possible that as Chinese people consume more foreign news, they become more informed of the democratic systems and practices in other countries, and consequently hold the government and legal authorities to higher standards. In other words, it is the strong contrast between the political environment in foreign countries and China that leads Chinese respondents who know more about legal and criminal justice principles and practices around the world to be more critical of the Chinese authorities. On the other hand, since Taiwan has developed into a democratic country and valued the protection of human rights, exposure to foreign news does not have a similar significant effect as in China.

Together, this evidence suggests a limited effect of the media on public trust in legal authorities in Chinese societies, supporting the cognitive response theory (Greenwald 1968; Perloff and Brock 1980). The theory argues that people are not passive recipients of persuasive information. Message processing is a very active and interactive process. Sometimes people are persuaded without much thought, but more often than not, they analyze the information received, compare information from different sources, and connect persuasive messages to their existing experience and feelings. If there are negative preexisting thoughts about legal authorities due to personal or vicarious experience, media propaganda cannot easily change such thoughts and may even create more antagonistic or cynical sentiments. Relatedly, people more trustful of the media should be more receptive to, and influenced by, persuasive messages in the media, and consequently, have greater trust in legal institutions, which is exactly what this study has found.

Explorative in nature, this study has several limitations. The cross-sectional nature of the data hinders the unraveling of causal relationships. Meanwhile, despite the use of random sampling techniques, migrant workers in China are likely to be underrepresented in a household sample due to their absence from the city household registration (hukou) and their high residential mobility.

In addition, public trust in police and courts are measured by two single items. Although these variables appear to have good validity, they may not adequately capture the complicated nature of the concept of trust (Stoutland 2001). Recent research on trust in police has distinguished conceptually and/or analytically between institutional- and motive-based trust (Hinds and Murphy 2007; Tyler 2005), trust and legitimacy (Hawdon 2008), trust and confidence (Jackson and Bradford 2010), trust and satisfaction (Sun et al. 2013), and perceptions of effectiveness, procedural fairness, and shared values/priorities (Hohl et al. 2010). Future research should incorporate more carefully designed survey items to better measure the various aspects of law enforcement and judicial performance.

Finally, although this study takes the initiate to explore the effects of the type of media, exposure to foreign news, and trust in media on trust in legal authorities in China and Taiwan, it does not examine the content of media coverage, particularly crime- and criminal justice-related materials, in China and Taiwan which may have a direct impact on public perceptions of legal authorities. Future research should conduct content analysis, documenting the amount and nature of media coverage allocated to police and courts topics and analyzing both manifest and latent content of these messages (Gray and Densten 1998). It is important that researchers not only quantify the links between media use and public trust, but also develop qualitative cultural understanding of the communicative roles that the media play in shaping public perceptions of criminal justice.