Introduction

Political trust, defined as a basic evaluative orientation toward public institutions and agents [16, 32], relates not only to the maintenance of regime stability but also to the nourishment of social coherence [6, 27, 38, 40, 58]. As such, it has always been on the research agenda of scholars who are interested in Chinese politics, prompting a large number of empirical studies (for some recent studies, see [6, 8, 54, 56].

Despite ample evidence of the trust level of specific political trusting targets accumulated in the current literature, a typology of political trustors that highlights distinct aspects of political support has not been well examined. By definition, this kind of typology refers to the classification of citizens (the trustors) according to their unique pattern of political attitudes. As an important theoretical and analytical tool for the social sciences [31], a typology of political trustors in China can further our understanding of Chinese politics from at least two perspectives. First, such a typology works better than a single index of political trust to clarify the theoretical meaning of political trust [17]. Indeed, previous studies on political trust have tended to focus on confidence in some specific political object (e.g., the central government, as in [42] or an omnibus but single measure that is derived from multiple items of political trust (i.e., the average trust of multiple institutions, as in [23]). Notwithstanding the insights gained from these studies, the estimated level of a single index of trust is insufficient to inform us the extent to which it can be understood to reflect a generalized type of political support undergirded by an unconditional and stable “faith” in the political system, or to reflect a specific type of political support that is contingent on the performance of institutional operations. This distinction between diffuse and specific political supports, as asserted by [10, 11], has practical importance since a low status of political trust would not be a critical concern for the whole regime if the meaning of political trust gravitates toward the diffuse instead of the specific type of support.

Second, a typology of political trustors helps to reveal the unique effect of the determinants of political trust. In the current literature, scholars generally adopt two fundamental approaches to accounting for the observed patterns of political trust: one is the institutional approach, which emphasizes perceived performance [44, 45], and the other is the cultural approach, which directs scholars’ attention to the hierarchical and authoritarian political culture in China [34, 39, 49]. However, these two theoretical approaches have not been well distinguished from each other empirically because the effects on political trust of their key variables—perceived institutional performance and adherence to authoritarian culture—always coincide with each other to be positive [37, 41, 51]. As a result, understanding where these two determinants of political trust in China can function differently is elusive. A typology, in this regard, provides such an opportunity to investigate this question, i.e., to see whether and how perceived performance and authoritarian culture have unique, or even contrasting, links to different types of trustors.

Against this background, this article draws on and extends the typology of political trustors proposed by Wu and Wilkes [49]. Specifically, we construct explicit measures for diffuse and specific political support, based on which we construct a four-way classification scheme of political trustors—compliants, critical citizens, contingentists, and cynics—for political institutions and actors. Using this typology, we investigate (1) the basic population distribution of each type of political trustors, (2) to what extent trust in political institutions maps onto or deviates from trust in political actors; and (3) how different types of political trustors are associated with perceived political performance and authoritarian culture.

Theoretical background

Diffuse-Versus-Specific Political Support

In the current literature, the concept of political trust has often been loosely defined as a basic evaluative or affective orientation toward public institutions and their employees (e.g., [32]. As such, the trust reported in empirical studies may concern a wide range of political objects [21, 22]. To clarify the objects of the trusting attitude, scholars widely cite Easton’s theory and distinguish between diffuse political support and specific political support [10, 11].

Specifically, diffuse support concerns people's support of “the regime as a whole and the political community” ([11], p. 445), so it reflects "generalized attitudes toward the system” [1, p. 63]. In contrast to diffuse support, specific support is mostly concerned with the performance of specific institutions. As such, a high level of specific support means that people are satisfied with how various specific political institutions or their staff are working in reality, which, by definition,Footnote 1 could differ from the generalized faithfulness in the overall political system or agents.

The distinction between diffuse and specific support has been widely cited, and it comes as no surprise that this dichotomy can be integrated into the process of constructing the typology of political trustors. One attempt in this regard comes from Wu and Wilkes [49].

A Previous Three-Way Typology Scheme

Wu and Wilkes [49] propose a three-way typology of political trustors. The first type is called compliants, who always trust all kinds of political institutions. The second one is named cynics, who can be viewed as the opposite of compliants in that they distrust all public institutions. Compliants and cynics, according to Wu and Wilkes [49], stand for political trustors who illustrate a diffuse type of support. In contrast, the third type of political trustor, called critical trustors, is believed to capture specific political support because people of this type have varied levels of trust across different political institutions, so their trust is bestowed to different political institutions not indiscriminately, but differentially according to which specific institution is concerned.

The original three categories proposed by Wu and Wilkes [49] are enlightening for research on political trust in China. However, there is still room for further improvement. First, although the Wu–Wilkes typology was proposed with the intention of accommodating the theoretical distinction between diffuse and specific support, the diffuse-versus-specific distinction is nevertheless not explicitly addressed and deployed. As a result, the meaning of the derived types of political trustors can be vague and called into question. For example, according to Norris [35] and many other scholars (e.g., [9], critical citizens are characterized by supportive attitudes toward the generalized political regime while not being very satisfied with how the political system is operating in reality. Following this line of thinking, critical citizens should reasonably feature a higher level of diffuse support and a low level of specific support. However, the critical trustors in the Wu–Wilkes typology who are defined in terms of response variations fail to show this nuanced combination of diffuse/specific support, a result that can be attributed to the lack of explicit measures of the two types of political support. It is also farfetched to believe that critical trustors are solely reflected by specific support.

Second, the standard of classifying citizens into different categories of political trustors as deployed in the Wu–Wilkes typology is mechanical and sometimes paradoxical. Compliants and cynics are a priori assumed to show no variation (zero standard deviation) in their attitudes toward different political institutions. This operationalization can be too strict. Moreover, this mechanical classification would inevitably suggest an unrealistic situation where the determinants of political trust, such as perceived performance and cultural values, either have the same effect across all kinds of political institutions or are simply independent from one's likelihood of being a compliant or a cynic. The operationalization of critical trustors in the Wu–Wilkes typology also induces paradoxical interpretations. Critical trustors, following the theory of critical citizens, should not be very satisfied with specific political performance, but the Wu–Wilkes typology does not exclude from critical trustors those who lean in the direction of having positive (although not the same) attitudes toward all political objects. In this case, some of the critical trustors cannot be “critical” at all.Footnote 2

Third, the objects of political trust in the Wu–Wilkes typology partially gravitate toward political institutions, but in addition to political institutions, we should also take into account political actors. Here, we do not use the term politician or authorities but follow Pippa Norris to adopt the term “actor” to emphasize the generic staff instead of solely the “heads” of public institutions [35]. This institution-versus-actor distinction has a long history in political research. For instance, in accounting for the decline in political support in the 1960s and the 1970s in the United States, Miller stresses the detrimental consequences of institutional policy alternatives (institutions), while Citrin attributes the major reason to the dissatisfaction in incumbent leaders (actors) [7, 32, 33]. In the Chinese social context, citizens’ attitudes toward public employees could also differ from those toward institutions. For example, the institutional arrangement could be appreciated by citizens, while those who execute that arrangement could be blamed for their misconduct or low efficiency [27]. Conversely, people might have strong confidence in political actors due to social connections and familiarities in local societies, but may not have faith in public institutions given that the administrative process can be alien and distant [3, 43, 53]. Regardless of which is the case, the institution-versus-actor distinction should be considered.

A Refined Four-Way Typology Scheme

The key to overcoming the shortcomings discussed earlier is to construct explicit measures for specific and diffuse support for political institutions and actors. Different types of political trustors can then be constructed by cross-tabulating these two types of political support. Instead of the three-way Wu–Wilkes typology, doing so would result in a four-way classification, as presented in Fig. 1.

Fig. 1
figure 1

A typology of political trustors

Full size image

Like the Wu–Wilkes typology, compliants are defined as those who always trust, but with the extension that this kind of persistent trusting attitude is further specified to apply to both diffuse and specific supports. In other words, compliants are those who speak highly of both the system as a whole and the works of individual institutions and actors. The same theoretical reasoning can be applied to cynics. Since they are always distrusting, there are good reasons to expect them to have a low-level value for both specific and diffuse support. Critical citizens, as discussed earlier, are supportive of the generalized political regime as a whole but not very satisfied with the way political institutions or actors work in reality. Following this line of thinking, we define critical citizens to be those who have a high value of diffuse support, but a low value of specific support. The last typology is called contingentists because their (perhaps high) political trust is driven by the extent of satisfaction with political performance that is contingent on specific circumstances. Relatively speaking, they are less likely to be affected by generalized faith in the regime, so their degree of diffuse support can be low.

Regarding this four-way typology scheme, two caveats should be addressed. First, the idea of critical citizens was proposed in the social context of Western democracies in the first place, so it is not our intent to mechanically apply the original conceptualization to the context of Chinese society. Instead, we draw on the theoretical merits of this idea by highlighting the potential discrepancies between diffuse and specific type of political support. This discrepancy has been noted to exist and matter for the political process in China, despite the powerful ideology and party-state system (e.g., [15, 28]. In this regard, the question at issue should be the formative mechanism of this kind of discrepancy, and this study attempts to approach this question by bridging different types of political trustors with the perceived political performance and cultural values.

Second, the contingentists constitute a genre of political trustors that has not been fully examined in previous literatures, but their basic information presented in this article bears significance in the Chinese society. Specifically, those contingentists, as characterized by their satisfaction with political performances while not being very much confident in the underlying order of political life, are kind of rational and secular in terms of their political attitudes, which comes as no surprise against the background of the market-oriented transition. Moreover, the existence of contingentists suggests that the principle of performancism that has been used to explain the tournaments among officials [26, 29] could be expanded to include ordinary citizens when they evaluate the extent of trustworthiness of the authorities. In this regard, the stability of the regime might be undermined due to the contingencies of such political trustors.

All four types of political trustors are constructed for both political institutions and actors. That is to say, for political institutions and actors, both diffuse and specific supports can be revealed. This deserves more discussions because previous studies often make a correspondence of diffuse political support to political systems, and a correspondence of specific political support to institutions and actors (e.g., [36]. This kind of correspondence is theoretically understandable, but we argue that it is inappropriate to interpret such a correspondence in a rigid fashion. The major reason is that the political system cannot be fully separated from either its institutions or actors. Instead, attitudes toward the general (rather than situational and particular) institutions and actors could shed light on one’s diffuse support for the system. Due to this consideration, we propose the operationalization of diffuse support for both institutions and actors.

Can We Find a Polarized Spectrum of Political Trustors?

The four types of political trustors are constructed based on the theoretical dynamics between diffuse and specific support, and their frequencies are an empirical question. In light of the limited research on the typology of political trustors, the current literature cannot direct us toward an elaborate hypothesis about the detailed percentages of different types. The original research of Wu and Wilkes [49] identifies approximately 35–40% of China’s population as compliants, 1–2% as cynics, and 59–63% as critical trustors, but as discussed earlier, these estimates are not based on explicit measures of political support, and different types of trustors can be confounded with each other (e.g., the critical trustors may include both critical citizens and contingentists). Hence, with regard to our four-way classification scheme, the empirical distribution is still an exploratory question.

Despite the exploratory nature of this research, we would like to argue here that it is very unlikely that a polarized pattern, i.e., a high percentage of cynics and compliants at the two extremes with both critical trustors and contingentists between then, will be observed. Political polarization in the United States has been a focus for social scientists over the past several decades, but until now, direct evidence to support it has been rare [13]. Among the various factors that might lead to political polarization of the masses, one that can be relevant for the case of mainland China is the socioeconomic polarization, such as concerted social inequality. However, unlike in the United States and many other societies, social inequality, based on previous studies, has a limited impact on people’s attitudinal polarization [47, 48]. Even for marginalized citizens such as rural residents, social inequality is often taken to be some minor and inevitable “side effect” of social development, so general attitudes toward the regime or its agents still lean toward the positive side [50]. In this regard, we suspect that compliants may occupy a considerable proportion of the respondents and that cynics cannot commensurately represent the other pole.

For the other two types of political trustors–critical trustors and contingentists, it is difficult to specify their relative percentages, but they both deserve more attention since a certain distinction between the regime as a whole and the specific institutions and their employees has always been in people’s minds (e.g., the distinction between country government and village cadre, as in [4]. In this regard, showing the extent of their current prevalence in China adds new information to the current literature.

Characterizing the Typology from Institutional and Cultural Accounts

Based on the refined four-way typology, we can further examine how perceived government performance and adherence to authoritarian culture, two widely noted factors related to political trust in China, can be related to the likelihood of belonging to a specific type of political trustor.

Specifically, the institutional approach to accounting for political trust argues that people have faith in public institutions that provide satisfactory performance. In other words, political trust depends on political performance, and citizens “rationally” evaluate such performance and then make judgments accordingly [34, 52]. This line of thinking has been used to explain the high level of trust in the Chinese government. The very high rate of China’s economic growth over the past several decades has been used to affirm the outstanding performance of the government. With the widespread improvement of living conditions, Chinese citizens, even marginalized citizens, have been noted to show strong faith in the government [12, 14, 48, 50, 55].

Relative to the institutional approach, the cultural approach shifts attention to the unique cultural political norms in China. Specifically, many scholars draw on traditional Confucian political thoughts, and emphasize the authoritarian culture in China that gives rise to a high level of political trust [37, 39]. The basic idea of this authoritarian culture is that the relationship between public institutions and citizens is not reciprocal but hierarchical, mimicking the parent–child relationship [2, 5, 24, 30]. On the one hand, the government, as well as many other public institutions, is viewed as benevolent to the people and committed to maintaining and promoting their welfare, similar to what parents do for their children. On the other hand, ordinary people receive care from public institutions, so they assume a subordinate status, similar to filial piety from children to parents. Clearly, according to this cultural logic, the high level of political trust in China is partly explained by the moral virtues attached to public institutions or actors.

Discussion of the institutional and cultural approaches suggests that perceived performance and the subscription to authoritarian culture are both trust promoting, and the current literature suggests no reason to suspect that this kind of positive effect has to be restricted to one particular kind of political support. That is, perceived outstanding political performance can improve one’s evaluation of institutions or actors (specific political support), as well as one’s confidence in the regime as a whole (diffuse political support). Similarly, culturally viewing governments or public employees as benevolent should also give rise to positive attitudes toward different individual political trustees (specific political support) and the overall political regime (diffuse political support). Altogether, compliants, due to their strong specific and diffuse support, could be characterized by better perceived performance and higher odds of embracing authoritarian culture. Conversely, the limited specific and diffuse political support of cynics would indicate that they could reveal a low level of perceived performance and authoritarian culture adherence. To summarize these discussions, we propose the following hypotheses:

H1:

Compliants, relative to other types, are more likely to have positive perceived political performance and a higher likelihood of adhering to authoritarian culture.

H2:

Cynics, relative to other types, are more likely to have limited perceived political performance and a lower likelihood of adhering to authoritarian culture.

The situation of critical citizens can be more complicated because perceived performance and authoritarian culture can have differential associations with the two types of political support. Specifically, as shown in Fig. 2, perceived performance is a theoretical construct that is geared toward the practical performance of different institutions, so this variable has an innate and stronger connection with specific support than diffuse support. In other words, if one perceives political performance to be good, his or her specific support should be raised to a greater extent than diffuse support. Relative to perceived performance, adherence to authoritarian culture is less contingent on situational peculiarities. Like other kinds of cultural norms, authoritarian culture should be taken as a stable and deep-seated propensity. As such, it should be more tightly related to diffuse support than specific support. Since critical citizens are conceptualized as those with a high level of diffuse support and a low level of specific support, reasonably, we would expect that among critical citizens, perceived performance is not that strong, while the likelihood of embracing authoritarian culture is high. The situation of contingentists should be reversed.

Fig. 2
figure 2

Illustrations of hypotheses for critical citizens

Full size image

Following this discussion, we propose the following hypotheses:

H3:

Critical citizens, relative to other types, are more likely to have limited perceived political performance and a higher likelihood of adhering to authoritarian culture.

H4:

Contingentists, relative to other types, are more likely to have strong perceived political performance and a lower likelihood of adhering to authoritarian culture.

Methodology

Sample

In this study, we take advantage of the Chinese Social Trust Survey (CSTS) to test the hypotheses listed above. The CSTS, collected in 2011, adopted a multistage random sampling strategy for urban and rural areas. In urban areas, the primary sampling unit is the city, and the secondary sampling unit is the neighborhood committee. In each sampled neighborhood committee, one residential community is sampled, and all residents are interviewed. In rural areas, administrative counties constitute the primary sampling unit, and the secondary sampling unit is the town. One village is randomly selected from each town, with all residents included for the subsequent interview. In total, the CSTS includes 3138 urban cases and 2158 rural cases, covering Nanjing (both urban and rural cases), Tianjin (both urban and rural cases), Chongqing (both urban and rural cases), Lanzhou (both urban and rural cases), Shenzhen (only urban cases), and Xi’ning (only rural cases). The response rate is 94.23% in urban areas and 98.09% in rural areas.

We use the CSTS instead of other large-scale surveys, such as the Asian Barometer Survey, World Values Survey, and the Chinese General Social Survey, because the CSTS includes very rich survey items regarding the two types of political support, perceived institutional performance and authoritarian cultural value adherence, and many other necessary control variables [57]. Many of these variables call for a battery of items that span multiple public institutions and actors, and they are not included in the other nationally representative surveys. Because of this unique merit, although the CSTS is not very up-to-date and the survey sites are limited,Footnote 3 it is still the ideal survey data source for our research questions.

Measures

Dependent Variables

The dependent variable of this article is the four-way typology, which is constructed based on the responses to diffuse and specific support.

Specific Support of Political Institutions

This variable is measured by the summation of the extent of respondents’ reported trust in the following institutions (1 = highly distrust; 2 = distrust; 3 = neutral; 4 = trust; 5 = highly trust): the central government, the provincial government, the municipal government, the district government (urban only), the sub district office (urban only), the county government (rural only), and the township government (rural only). The Cronbach’s alpha for these items for the urban sample is 0.93, and for the rural sample, it is 0.90.

Specific Support of Political Actors

This variable is measured by the summation of the extent of respondents’ reported trust in the following political actors (1 = highly distrust; 2 = distrust; 3 = neutral; 4 = trust; 5 = highly trust): government staff (rural only), civil servants (urban only), urban management and law enforcement personnel (urban only), judges, policemen/women, and soldiers. The Cronbach’s alpha for these items for the urban sample is 0.85, and for the rural sample, it is 0.81.

Diffuse Support of Political Institutions

This variable is measured by the summation of the extent of respondents’ agreement with the following statements (1 = strongly disagree; 2 = disagree; 3 = neutral; 4 = agree; 5 = strongly agree): “Policies formulated by the government can represent the interests of the majority of people,” “Governments’ policies are stable,” “The government's work report is credible,” “The government's price hearing is just a formality (only for the urban sample, the order of option is reversed for this item),” “The government's fiscal expenditure is in accordance with the government's fiscal budget and effective,” and “The selection of cadres is fair (only for the rural sample).” The Cronbach’s alpha for these items for the urban sample is 0.54,Footnote 4 and for the rural sample, it is 0.720.

Diffuse Support of Political Actors

This variable is measured by the summation of the extent of respondents’ agreement with the following statements (1 = strongly disagree; 2 = disagree; 3 = neutral; 4 = agree; 5 = strongly agree): “Most government officials are public servants of the people,” “Government officials often only care about their own political achievements (the order of option is reversed for this item),” “Law enforcement officers do not operate according to law (the order of option is reversed for this item),” and “Most officials are embezzlers and corrupt (the order of option is reversed for this item).” The Cronbach’s alpha for these items for the urban sample is 0.68, and for the rural sample, it is 0.66.

Based on the reported diffuse and specific supports, we can construct our dependent variables as described earlier. Specifically, compliants refer to those whose reported diffuse and specific support are larger than their respective median values. Cynics are operationalized as those whose reported diffuse and specific support are smaller than their respective median values. Critical citizens are operationalized to be those whose diffuse support is larger than its median while specific support is smaller than its median. Contingentists, similarly, are operationalized to be those respondents whose diffuse support is smaller than the median while specific support is larger than the median.

The typology is constructed respectively for political institutions and political actors.

Independent Variables

The key predictors in this article are perceived institutional performance and authoritarian culture. For perceived institutional performance, the CSTS measures of the extent of people’s satisfaction with the performance of public institutions in terms of food security, disaster relief, environmental protection, compulsory education, employment services (urban only), philanthropic services, legal enforcement, household registration, pension services, health care, social security, and affordable housing construction (urban only). The options are coded to be 1 = highly unsatisfactory; 2 = unsatisfactory; 3 = neutral; 4 = satisfactory; 5 = highly satisfactory. The responses to these items are summed. The Cronbach’s alpha for these items for the urban sample is 0.92, and for the rural sample, it is 0.86.

Authoritarian culture is measured by attitudes toward the following three statements: (1) In general, what government officials do is always right (faultless); (2) The head of the government resembles the head of a family, so we should always respect their decisions (family head); and (3) The government can decide what can be discussed in our daily lives (opinion control). The options are 1 = totally disagree; 2 = disagree; 3 = neutral; 4 = agree; and 5 = totally agree. We compute the average value of the answers to these items. The Cronbach’s alpha for these items for the urban sample is 0.79, and for the rural sample, it is 0.76.

Control Variables

Many covariates are controlled for in the following analyses, including: gender (1 = female; 0 = male), age, age square, marital status (1 = married; 0 = otherwise), ethnicity (1 = Han; 0 = minorities), household registration (hukou) status (1 = agricultural; 0 = otherwise), CCP membership (1 = yes; 0 = no), log-transformed annual income, and educational attainment (1 = junior high school and below; 2 = senior high school; 3 = college and above).

Among these variables, gender, age, marital status, ethnicity, educational attainment, and log-transformed annual income refer to basic sociodemographic and socioeconomic information, so they should be fixed. Household registration (hukou) status represents the institutionalized (dis)advantages in China, which could have a direct effect on people’s political trust. CCP membership is a kind of political identity that is straightforwardly related to the confidence in public institutions and their staff.

Detailed descriptive statistics are reported in Appendix Table 2.

Analytical Strategies

Based on the typology of political trustors, we use descriptive statistics to depict the general situation of compliants, cynics, contingentists, and critical citizens for institution-based and actor-based trust, respectively. The logistic regression model is adopted to see how perceived performance and authoritarian culture are related to different types of political trustors.Footnote 5 Because the CSTS collected information in urban and rural areas, we perform the analyses separately for the urban and rural samples.

Results

Empirical Patterns of the Political Trustor Typology

We first examine the distributional characteristics of the different constructed types of political support. To enhance our comparison, we standardize these variables. As shown in Fig. 3, in urban and rural China, both institution-based and actor-based diffuse support lean toward the right side (left skewed), which is also the case for specific support. This distributional character implies that, on average, the surveyed respondents have a generally positive attitude toward public institutions and political actors, both in the diffuse and the specific senses. Based on the two estimated types of support, we construct the typology of political trustors, with each type's percentage shown in Fig. 4. In both urban and rural areas and for both institution-based and actor-based trust, we observe the largest proportion of respondents can be classified as compliants, which may be as high as approximately 50% across geographical regions and trust targets (institutions and actors). Relative to compliants, the other three types of political trustors are each estimated to comprise smaller segments of the population: between 10 and 20%. Taken together, using the typology of political trustors constructed based on the dynamics between diffuse and specific support, we detect a preponderance of compliants, although cynics, critical citizens, and contingentists also represent a considerable fraction of the population.

Fig. 3
figure 3

Empirical distributions of different types of political support

Full size image
Fig. 4
figure 4

The descriptive information of different types of political trustors

Full size image

Is Trust in Institutions Always Consistent with Trust in Actors?

With the constructed typology, readers might be interested in the degree of correlation between trust in political institutions and trust in political actors across different types of trustors. To answer this question, we cross-tabulate institution-based and actor-based political trust, as shown in Table 1.

Table 1 The cross tabulation between political institutions and actors
Full size table

Despite the general positive relationship (the odds ratios are all larger than unity and significant at the 0.001 level), there is still a noticeable proportion of respondents whose trust in political institutions differs from their trust in political actors. The percentage of discrepancies between political actors and institutions varies from 16.45 to 24.97% in urban China and from 11.43 to 20.38% in rural China. In light of these findings, despite the general positive correlation, one’s confidence in political institutions does not always suggest confidence in political actors. This finding lends support to the necessity of making a distinction between political institutions and actors.

The Institutional and Cultural Accounts

In this subsection, we investigate the relevance of perceived government performance and authoritarian culture. The analytical results of the logistic regression models are presented in Fig. 5 (detailed results can be found in the Appendix Table 3.

Fig. 5
figure 5

Analytical results of the relevance of the authoritarian culture and perceived government performance

Full size image

Clearly, both perceived performance and authoritarian culture are positively correlated with the odds of being a compliant and negatively correlated with the odds of being a cynic. This pattern is affirmed in both urban and rural China. Hence, both H1 and H2 are supported. In contrast to these two types of political trustors, critical citizens, in general, have a significantly lower value for perceived performance but a higher value for authoritarian culture. The patterns of correlations of these two factors are simply reversed for contingentists. In this light, both H3 and H4 are affirmed.

Concluding Remarks

Political trust has been on the research agenda of social scientists who are interested in China. However, the operationalization of political trust in many previous studies is somewhat partial, focusing on the perceived trustworthiness of a particular institution, reported trust in some specific politicians, or an average level of trust across multiple public institutions. The more refined institution-versus-actor distinction and the diffuse-versus-specific distinction are not well illustrated. This article draws on explicit measures of specific and diffuse political support and extends this line of scholarship by presenting a refined typology of political trustors. For political institutions and political actors, we distinguish between compliants who always trust specifically and diffusely; cynics, who always distrust specifically and diffusely; critical citizens, whose diffuse support is high while specific support is low; and contingentists who have a high level of specific support and a low level of diffuse support.

Using data from the Chinese Social Trust Survey 2011, we find that, overall, both specific and diffuse political support lean toward the positive end. With regard to both institution-based and actor-based political trust, compliants comprise the largest proportion. Relatively speaking, the numbers of the other three types of political trustors are smaller. This study also highlights the potential difference between trust in institutions and trust in political actors, since they could diverge from each other despite a general positive correlation. The average level of perceived institutional performance and adherence to traditional authoritarian culture are both positively correlated with the likelihood of being a compliant, and both are negatively correlated with the likelihood of being a cynic. However, authoritarian culture promotes the likelihood of being a critical citizen, while perceived institutional performance abates that likelihood. The pattern for contingentists is reversed.

This study provides further empirical evidence for the prevalence of compliants in Chinese society [19], a finding that is consistent with Wu and Wilkes [49] despite their different measurement scheme. Compared with previous studies that document a strong confidence level in some particular trust targets, this study suggests that this strong confidence can be both diffuse and specific. With regard to this finding, we are aware of potential methodological concerns, such as the social desirability bias,Footnote 6 but we are still inclined to view it as some evidence for the generally supportive attitudes of Chinese citizens toward the regime as well as its constituent components [18, 20, 21].

Relatedly, one concern that is often mentioned in the scholarship of Chinese political trust research is the measurement validity of the trust items. That is, scholars often attribute the reported high level of trust to the factor of political fear rather than some unbiased “true” attitudes. This article joins in this strand of scholarship by showing that the level of reported trust is also related to the type of trustees (targets of trust) at issue. Therefore, when evaluating the validity of political trust measures in China, scholars should specify clearly what kind of trustees are concerned.

The critical citizens documented in this article deserve more attention [46]. Readers should keep in mind that critical citizens are not cynics in that their diffuse support for the whole regime is still strong. In addition, further supplementary analyses suggest that not every political institution and actor is criticized by them. The targets of their critique are mostly local governments and some specific political actors (e.g., urban management and law enforcement personnel in urban areas). In this regard, the existence of these critical citizens should not be deemed a source of crisis for this regime. Instead, they may stand for a social force that contributes to regime “health” since critical citizens can encourage reforms that improve the accountability of institutions [35].

Through the nuanced typology of political trustors, this article shows a situation where perceived performance and authoritarian culture can work in different directions. Although both can promote (reduce) the likelihood of being a compliant (cynic), their roles in the cultivation of critical citizens and contingentists differ. The diffuse political support of critical citizens can be enhanced by the embracement of authoritarian culture, but critical citizens are noted to have a perception of limited government performance. The case of contingentists is reversed. This unique combination in the effects of perceived performance and authoritarian culture suggests the potential for a transition from one type of political trustor to another. For instance, further improvement of perceived performance can help convert critical citizens toward compliants, while the propaganda of authoritarian culture might “push” contingentists toward compliants.

This article also provides insights for the comparative political research. For instance, previous studies often compare the substantive evaluations of survey items from one society to another (e.g., the extent of trust in the central government), but this article shows that in addition to the substantive responses to specific items, cross-national comparative political studies can also compare the “structural aspect” of the items, which are embodied by the different types of political trustors. Related to this point of view, the comparative political researchers should always pay attention to the exact meaning of some pre-defined political categories. For instance, the way to define and operationalize critical citizens might differ between China and other countries. In this case, direct comparison of empirical results would be called into question, since seemingly cross-national variations might be driven by the different conceptualizations.

This study characterizes the different types of political trustors with regard to the two major explanatory factors of political trust in China: perceived performance and authoritarian culture. However, other than these, the multivariate analyses do not show noticeable sociodemographic patterns. That is, the basic sociodemographic characteristics do not vary notably from one type of political trustor to another. In this regard, more explanatory variables should be collected to further our understanding of the formation of the typology of political trustors.

Several limitations of this study should be discussed. The foremost limitation is that with the current data set, we cannot rule out the potential issue of social desirability bias. In the social context of China, as reflected in many previous articles, this can be tantamount to the inflation of reported political trust. The other limitation lies in the shortcoming of the measurement. For instance, authoritarian culture can be expanded to refer to a wider range of political attitudes or even nonpolitical attitudes. Finally, the use of the cross-sectional data means that we cannot ascertain a strict causal relationship. Bearing these limitations in mind, we would like to view this article as exploratory, and our objective is to draw scholars’ attention to the multiple dimensions of the concept of political trust in contemporary China.