Better support for national than local system during the COVID-19 pandemic in China.

Individuals increase their support for social systems in response to the threat, panic, and uncertainty that characterized the COVID-19 pandemic. This could be because a powerful social system can compensate for a lack of control at the individual level. However, the levels of public support for national versus local systems could be different in China. Two studies investigate whether people support the national more strongly than the local system during the COVID-19 pandemic. Study 1 analyzed data of 3593 participants from China; the results showed that participants reported higher levels of support for the national system than the local. In Study 2, we further tested a possible moderator for it. With a sample of 275 participants, we found that the difference between public support for national and local systems in China was based on the perceived higher response efficacy with the national government. Implications for research on system justification and governmental pandemic responses were discussed.

The Coronavirus disease 19 (COVID‐19) pandemic has threatened people's physical health and psychological well‐being globally (Duan & Zhu, 2020). In addition to preventative behavior at the individual level, the societal system to which an individual belongs also greatly influences the individual's risk of contracting the virus (Wilkinson, 2020). In most countries, national and local governments have played and continue to play pivotal roles in responding to the threat of the virus, taking concrete measures to limit its spread (Ferguson et al., 2020). At the same time, members of the public respond to the government's performance in adapting to the pandemic (Gadarian et al., 2021). Public perceptions of governmental performance could have an impact on citizens’ concerns regarding the threats and risks posed by COVID‐19. That means if governments are trustworthy or taking effective actions, the public may suffer less psychological stress from the pandemic. Surveys proved that trust in the government was negatively correlated with risk perception (Ma & Christensen, 2019), and public fear and anxiety during severe acute respiratory syndrome (SARS; Cheung & Tse, 2008).

System justification theory (SJT; Jost & Banaji, 1994) suggests that people are likely to increase their support for systems during threatening situations (e.g., Day et al., 2011; Ullrich & Cohrs, 2007). Analysis from different countries also provides evidence for the “rally around the flag” effect, in which people increase support to political leaders facing international crises, during the COVID‐19 pandemic (e.g., Baekgaard et al., 2020; Yam et al., 2020). Most of the studies previously conducted around this topic focused on the attitudes of people toward the entire socio‐political system in various countries (e.g., Day et al., 2011; Jolley et al., 2017). However, public attitudes toward local and national governments could differ (Gupta et al., 2020) in terms of factors such as political trust (Li, 2004) and public policy satisfaction (Tang & Yu, 2015). System support, which refers to the diffuse support for any systemic aspects of the polity, is more closely related to individuals’ psychological needs and responses in times of crises (Intawan & Nicholson, 2018; Tan et al., 2016), compared with other political indicators such as political trust. Thus, this study aimed to investigate whether there was a difference in the degree of public support for national and local systems during the COVID‐19 pandemic in China and to test a possible prerequisite for this phenomenon.

THREAT AND SYSTEM SUPPORT

SJT states that people are motivated to justify, defend, and support the systems to which they belong (Jost, 2019; Jost & Banaji, 1994) and that this motivation is intensified when they are exposed to system threats (Friesen et al., 2019; Kay et al., 2009). We use the term “threat” to refer to natural and social events that can harm the stability and political support of the system, such as terrorist activity, natural disasters, and public health emergencies (Napier et al., 2006). Events that threaten system fairness and legitimacy could also endanger, at the individual level, relational, epistemic, and existential needs, which lead to psychological stress (Jost, 2019). Conversely, support for the existing societal system serves the palliative function of reducing uncertainty and emotional distress (Jost & Hunyady, 2002), increasing subjective well‐being and physical health (Napier et al., 2020). Therefore, individuals will likely increase system support to respond to threats, panic, and uncertainty during circumstances, such as the COVID‐19 pandemic.

Previous research under the umbrella of SJT has proven that exposure to threats can increase system justifying responses. For example, in experiments (Landau et al., 2004) conducted after the 9/11 attacks, support for President Bush among respondents was increased after responding to questions about death or 9/11. Further studies also discovered that compared with the control condition, participants who were reminded of terrorist attacks demonstrated more support for the system (Ullrich & Cohrs, 2007). Similarly, in the case of Hurricane Katrina, some people justified the system by victim‐blaming and internal attributions for their own misfortune (Napier et al., 2006). Even reading passages about criticism of the social system would increase participants’ support for the system (e.g., Day et al., 2011; Jolley et al., 2017). More importantly, the threat posed by the COVID‐19 outbreak has also been proved to trigger public systems justifying responses in different countries. For example, Yam et al. (2020) confirmed public support for national leaders in 11 countries (Australia, Brazil, Canada, France, Germany, Hong Kong, India, Japan, Mexico, the United Kingdom, and the United States) increased in the early months of the COVID‐19 pandemic.

DIFFERENCE BETWEEN NATIONAL AND LOCAL SYSTEM SUPPORT

Although people would increase system‐justifying responses when the system was under threat, most available research focuses on whole political or economic systems. However, individuals’ motivation to justify different elements of the status quo could be different (Day et al., 2011; Laurin et al., 2011). In most countries, national and local systems are both essential actors during emergencies, such as pandemics (Gupta et al., 2020), and are responsible for activities, such as announcing information and taking preventative measures. However, they play different roles and sometimes there are conflicts between local and national respondents (Chambers et al., 2012). The national government coordinates the responses they receive from their people and synthesize the expertise of national agencies, while local governments may make more region‐specific policies and actions to contain pandemic outbreaks (Berman, 2020). People may hold different perceptions and attitudes toward national and local systems. Thus, in the present study, we focused on the difference between the public's support for national and local systems. Although there are no empirical studies that have directly tested this hypothesis, some research findings have provided evidence for the diversity of system support. For example, Kay et al. (2009) found that university students were more likely to support the federal government's funding of policies than that of their own university when they read passages about how they depended on the federal government. In another study, high school students justified the American system more strongly than the popularity hierarchy and social network at their school after reading a passage describing American society as deteriorating (Wakslak et al., 2011).

Is local system support different from national system support when people are threatened by the COVID‐19 pandemic? Kay et al. (2008) proposed a compensatory control mechanism to explain the system justification tendency; that is, people would have recourse to external systems of control to restore perceived control to a baseline, when personal control is low (Kay et al., 2009b, 2010; Landau et al., 2015). Previous studies have found that the circumstances in which individuals experience uncertainty and powerlessness, such as the pandemic, increase individuals’ tendency to defend and support the external system of a higher power, especially the government (Landau et al., 2015). Thus, support for the national system, which is more powerful for pandemic control, could be better at providing compensatory control. We propose that public support for a more powerful national system might be higher than for local systems.

In addition, we expect this difference to be more prominent in China due to its socio‐cultural characteristics, for the tradition of Confucianism praises the merits of authority and blames failures on the flaws of subordinate officials (Yang & Tang, 2010). In this research, therefore, we tested the hypothesis that during the ongoing COVID‐19 pandemic, the national system gets higher public support than the local system in China (Hypothesis 1).

SYSTEM SUPPORT AND GOVERNMENT PERFORMANCE

As mentioned earlier, individuals tend to support powerful external systems to maintain their perceived sense of control, and the national government, which tends to have more control and power, receives more public support than the local system. However, if the system does not provide effective structure after threats, individuals might not increase their support for this system (Friesen et al., 2014; Shepherd et al., 2011). Therefore, government compensatory control is based on structure and order (Kay et al., 2009), which could be reflected in government performance, and might affect public system support. For instance, an analysis of the World Values Survey data found that, compared with corrupt countries, people from less corrupt countries are more likely to agree that governments should take more responsibility for the public when they lack personal control (Kay et al., 2008). This relationship between the government and the public can be understood in terms of a response to the threat and systemic support; this was observed after the occurrence of natural disasters, such as Hurricane Katrina. The government's failure to respond to the natural disaster there raised queries about the legitimacy of those holding high‐level positions (Napier et al., 2006).

The efficacy of the government's response, which serves to provide a structure in an otherwise threatening situation, has been an important manifestation of governmental control during the COVID‐19 pandemic (Carter & May, 2020). Public support for a policy depends on the perceived effectiveness of that policy (Huber et al., 2020). More importantly, only if the pandemic response is more effective can the external system provide more control compensation to individuals otherwise lacking in control, thus obtaining higher public support. Thus, we suggest that the difference between the perceived efficacy of local and national governments might be a boundary condition differentiating national and local systemic support. Specifically, we predicted that the relationship between the system level (national system vs. local system) and system support was moderated by the difference in the perceived efficacy of governments’ responses (Hypothesis 2). Support for the national system could be higher than for the local system when the perceived effectiveness of the national government is higher than that of the local government. Support for the national system might be no different, or lower, from support for local systems when the perceived effectiveness of national government is lower than that of the local government.

CURRENT STUDIES

This research aims to examine the difference between public support for national and local systems. We conducted two studies during the COVID‐19 pandemic in China to test our hypothesis. In Study 1, we examined whether national system support was higher than local system support in a nationwide survey. In Study 2, we further tested the difference between system support for national and local systems, moderated by the difference in the perceived efficacy of governmental responses.

STUDY 1

Method

Both Study 1 and Study 2 were approved by the review board at the author's institution, and the questionnaire was completed voluntarily after participants gave informed written consent.

Participants

Study 1 was a national‐based survey conducted in China during the COVID‐19 pandemic. We recruited 5528 participants from the Tencent Questionnaire, an online platform in China that is similar to Amazon's Mechanical Turk. The final sample consists of 3631 participants (2024 men and 1607 women), aged between 17 and 61 years (M = 29.8, SD = 7.75). There were 1896 additional respondents but were excluded due to either failing the attention check (choosing a wrong option in instructed items) or having a response time less than 3 s per item. One participant who reported location as outside mainland China was also excluded. Of the total sample, 1562 completed the survey on January 31 or February 1, 2020 (Wave I), 933 on February 11 or 12, 2020 (Wave Ⅱ), and 1136 on March 24 or 25, 2020 (Wave III). The sample size represented the largest number of participants that have been recruited during the predetermined recruiting data phase. Participants were from 31 provinces, and 31.9% of them were from Hubei province, the center of the COVID‐19 pandemic with the highest number of infected patients1 during the first outbreak in China (Ainslie et al., 2020). Table 1 summarizes the broad sociodemographic patterns of the sample (see Supplemental Material for province details).

TABLE 1

Sociodemographic statistics for the sample of participants in Study 1

Sociodemographic characteristic	Valid percentages for the sample of participants (n = 3631)
Gender
Male	55.7%
Female	44.3%
Educational attainment
Higher level (at least regular college level)	58.0%
Lower level (up to regular college level)	42.0%
Age
17–30	63.6%
31–40	26.1%
Over 41	10.3%

Measures

Following previous research (Cichocka et al., 2018), three items from the general system justification scale developed by Kay and Jost (2003) were adapted to the context of the pandemic to measure national or local system support. Items were “During the COVID‐19 outbreak, I found China's (or local) society to be fair”; “During the COVID‐19 outbreak, China's (or local) political system operates as it should”; and “China's (or local) society needs to be radically restructured” (reverse‐scored). Participants rated each item on a 7‐point scale (1 = completely disagree, 7 = completely agree). The score was calculated for each participant by taking the mean of their responses to the three items (α= .68 and .66).

Results and discussion

We first used confirmatory factor analysis to test competing models that defined single or two (one for the local, one for the national level) factors to distinguish support for the national and local system. Five indicators, χ 2, standardized root mean square residual (SRMR), root mean square error of approximation (RMSEA), comparative fit index (CFI), and Tucker–Lewis index (TLI), were used to evaluate model fit. The two‐factor model exhibited acceptable fit2: χ 2 (7) = 225, p < .001, RMSEA = .093, 90% CI [.082, .103], SRMR = .018, CFI = .981, and TLI = .958. The single‐factor model did not fit the data well: χ 2 (8) = 918, p < .001, RMSEA = .177, 90% CI [.167, .187], SRMR = .036, CFI = .919, and TLI = .848. A comparison of the models reflected that the two‐factor model fits better than the single‐factor model, ∆χ 2 (1) = 693, p < .001. Thus, the six scales were employed with two factors: national system support and local system support.

The relevant descriptive statistics are shown in Table 2. We conducted a repeated‐measures analysis of variance (ANOVA), considering the wave as a between factor and the systemic level (national vs. local) as a within factor, to test the difference between national and local system support. Assumptions of repeated‐measures ANOVA analysis were tested. Normality assumption holds as indicated by skewness and kurtosis, in combination with Q–Q plot. It was found that all the skewness and kurtosis values of these repeated‐measures ANOVA tests were within the range from −1 to 1. According to the Q–Q plot, all the data points were well distributed along the diagonal line, indicating a normal distribution. In addition, the sphericity assumption was met as the repeated‐measures variables have only two levels, national system support versus local system support. Therefore, no correction was made in the analysis. As hypothesized, our analysis revealed a main effect across the levels of system examined, F(1,3628) = 417.90, p < .001, η 2 = .103, demonstrating that national system support was higher than local system support. The analysis also yielded a main effect of wave, F(2,3628) = 60.86, p < .001, η 2 = .032; system support increased with time, ps < .001 for each least significant difference test. There was no significant interaction between the levels of system and wave, F(2,3628) = .12, p = .891, η 2 < .001. We then included age, gender, educational attainment, and the respondent's province as covariates; the main effects remained significant after including the covariates. Thus, our hypothesis that the public's national system support was higher than their local system support in China during the pandemic was confirmed.

TABLE 2

Descriptive statistics of system support in Study 1

	Wave Ⅰ		Wave Ⅱ		Wave III		All
	M	SD	M	SD	M	SD	M	SD
National system support	5.40	1.24	5.57	1.15	5.87	1.02	5.59	1.17
Local system support	5.09	1.27	5.26	1.22	5.55	1.06	5.28	1.21

Considering the more serious and dangerous situation of Hubei, compared to other provinces during the initial outbreak of COVID‐19, we separately conducted paired samples t‐test (one‐tailed) for participants from Hubei and other provinces. For participants from the Hubei province, national system support was significantly higher than the local system support, t = 17.25, p < .001, d z = .507. For participants from other provinces, the difference is smaller but still significant, t = 13.47, p < .001, d z = .271. Since the results might be highly dependent based on the context of provinces, we further tested this effect in eight different regions in China. Except for the Northwest region, where we could not recruit enough representative samples and the difference was not significant, participants’ support for the national system was significantly higher than the local system in the other seven regions (see Supplemental Material for details). Figure 1 shows the effect size of all regions.

FIGURE 1

The difference between national and local system support in eight regions and Hubei province in China. The data of the Yangtze River region does not include Hubei province

STUDY 2

Participants

As suggested by Schönbrodt and Perugini (2013), a sample of 250 participants could achieve a stable estimate in a correlational study. Given the possible loss of participants, we recruited 280 participants from the Tencent Questionnaire (an online participant recruitment platform in China) on March 27 and 28, 2020. Five individuals were excluded from the study based on attention check, and the final sample was 275 participants (159 men and 116 women), aged between 16 and 46 years (M = 26.2, SD = 5.94). With this sample, the paired samples in the t‐test analysis had 80% power to detect an effect size of d z = .150 (one‐tailed), and the linear bivariate regression analysis had 80% power to detect an effect size of b = .11 (std dev σ_x = .994, std dev σ_y = .72). Table 3 summarizes the broad sociodemographic patterns of the sample.

TABLE 3

Sociodemographic statistics for the sample of participants in Study 2

Sociodemographic characteristic	Valid percentages for the sample of participants (n = 275)
Gender
Male	57.8%
Female	42.2%
Educational attainment
Higher level (at least regular college level)	65.1%
Lower level (up to regular college level)	34.9%
Age
16–30	82.5%
31–40	13.5%
Over 41	4.0%

Measures

Participants were asked to complete a survey about their social attitudes. The measure of national and local system support was slightly adapted from the 8‐item general system justification scale developed by Kay and Jost (2003; see also, Intawan & Nicholson, 2018; Wakslak et al., 2011). Sample items included: “China's national (or local) political system operates as it should,” “China's national (or local) society is getting worse every year” (reverse‐scored), and “China's national (or local) society is set up so that people usually get what they deserve.” Participants rated each item on a 7‐point scale (1 = completely disagree, 7 = completely agree). The score was calculated for each participant by taking the mean of their responses to the eight items (two of which were reverse‐scored, α = .86 and .88).

We assessed the perceived efficacy of local and national government's responses using two questionnaire items: “In general, I am satisfied with the prevention and control work of the Chinese (or local) government during the pandemic.” Participants responded on a 7‐point response scale ranging from 1 (completely disagree) to 7 (completely agree). The difference in score was calculated by subtracting the scores of the local government from those of the Chinese government. Thus, a higher difference in score indicated the perception of more efficacy in the national government's response, compared with that of the local government.

We have reported all relevant measures, data exclusions, and how we determined our sample size for Study 1 and Study 2. Data and materials are available at https://osf.io/qr4gp/?view_only=cd8c0960c94d444f9949b10345cfeedf

Results and discussion

Descriptive statistics are shown in Table 4. We conducted a paired samples t‐test to calculate the difference between national and local system support. Consistent with our proposal and Study 1 results, national system support was higher than the local system support, t = 5.98, p < .001, d z = .361. The paired samples t‐test also proved that the perceived efficacy of the national government's response (M = 6.16, SD = 1.01) was higher than the perceived efficacy of the local government's response (M = 5.90, SD = 1.22), t = 4.31, p < .001, d z = .260. Meanwhile, correlation analysis also indicated that the perceived efficacy of government responses was positively related to system support at both levels for the national system, r = .571, p < .001, and for the local system, r = .571, p < .001.

TABLE 4

Descriptive statistics for all variables in Study 2

		M	SD	1	2	3	4	5
1	Difference in perceived efficacy of governments’ response	.26	.99	—
2	National system support	5.36	1.10	.01	—
3	Local system support	5.10	1.13	−.17**	.79***	—
4	Age	26.16	5.9	−.08	.01	.09	—
5	Gender	—	—	.01	.03	.04	.19**	—

p < .01; *** p < .001. .

We conducted a moderation analysis using MEMORE macro for SPSS to test the moderating role of the difference in the perceived efficacy of the governments’ responses (Montoya, 2019). Bootstrapping was set to 5000 resamples. The result yielded an interaction effect between the difference score and the levels of systems (b = .20, 95% CI [.12, .28], SE = .04, t = 4.78 p < .001, R 2 = .077), indicating that the difference in the perceived efficacy of the governments’ responses significantly moderates the relationship between the levels of system and system support. Furthermore, as shown in Figure 2, when the difference score was higher (+1SD), the system levels (local and national) significantly predicted system support, b = .46, 95% CI [.34, .58], SE = .06, t = 7.77, p < .001, indicating that national system support was significantly higher than local system support. However, when the difference score was lower (−1SD), the system levels did not predict system support, b = .06, 95% CI [−.06, .18], SE = .06, t = 1.01, p = .314, indicating that national system support was not different from local system support. Thus, these results support our hypothesis that the national government's perceived response efficacy being higher than that of the local government would be a boundary condition of the difference between national and local system support.

FIGURE 2

The difference in the perceived efficacy of governments’ responses influences the relationship between the levels of systems and system support

GENERAL DISCUSSION

Across two survey studies, we examined the difference between public support for national and local systems in China and identified a moderator for the relationship between the levels of systems and system support during the COVID‐19 pandemic. We found that the Chinese public's support for the national system was higher than that for the local system. This difference is based on the premise that the public perceived a better performance from the national government than the local government, indicating that the national system had a greater degree of control in the pandemic and could thus be better than local systems at providing compensatory control to individuals.

Previous research on system justification has documented a difference in support for different systems (e.g., Kay et al., 2009; Wakslak et al., 2011). In line with these studies, we found that Chinese people show a higher degree of support for the national system, compared with the local system. This result is also consistent with studies examining differences in political trust, which yielded converging results, suggesting a rank differential pattern of “strong central, weak local” sentiment in China (Liu & Raine, 2016). This study suggests that a similar pattern shapes system support. Moreover, we tried to explain this phenomenon using a compensatory control mechanism (Kay et al., 2010) and further examined the moderation effect of the difference in the perceived efficacy of local and national governments’ responses during the COVID‐19 pandemic (Study 2). Where some researchers have argued that the responses of governments in catastrophes, such as Hurricane Katrina, have an effect on people's system‐justifying mechanisms (Napier et al., 2006), we provided empirical support for a positive relationship between government performance and system support during the pandemic. More importantly, the national government's higher perceived response efficacy may reflect greater control on the part of the national government as compared to support for the local system since the national government is better positioned to meet people's need for structure and mitigate their uncertainty and panic in this threatening situation.

There is another possible explanation for the differential levels of system support from a motivational perspective. Prior research has established that people will increase system justification motivation when exposed to system criticism (Cutright et al., 2011; Kay et al., 2005). Thus, criticisms of China from other countries may enhance Chinese system support. People more strongly justify the system being directly criticized (Kay et al., 2009; Wakslak et al., 2011), and thus the effect of national system criticism may differ between national and local systems. However, we found that participants from Hubei province (i.e., the center of the COVID‐19 pandemic in China), where the government had received more criticism from citizens, reported lower (not higher) support for the local system than others in Study 1 (see Supplemental Material for details). Therefore, the relationship between system criticism and system support might not operate simply in one direction as evidenced in this pandemic circumstance in China.

Even though we found the difference between national and local systems to be caused by compensation to people's personal loss of control under threat from the virus, in this study, we only examined our hypothesis in mainland China. In other countries, public system support might be dependent on other factors, such as the level of corruption (Anderson & Tverdova, 2003) and the ideology of the leading political party. Studies on political trust have found different patterns in East Asian countries and Western countries (Zhang et al., 2019). Meanwhile, we also found that the differences between national and local system support were not equal in eight regions in China in Study 1 (as shown in Figure 1). The Northern coast region, with a more prosperous economy, has a smaller effect size of difference than other regions; except for the Northwest region, we could not recruit a representative sample. Therefore, one must be careful and consider the cultural and contextual differences when generalizing our findings.

Our findings have several theoretical and practical implications. The present results expanded the SJT by highlighting a pattern of national and local system justification in China during a real circumstantial threat. Although there tend to be general motives under different system justifications (Jost, 2019), people may justify some systems more strongly than others (Wakslak et al., 2011). The result of the higher perceived effectiveness of the national government response as a boundary condition of the difference between system support also confirmed the compensatory control theory. Systems responding effectively, that is, exhibiting more control in threatening situations, tend to receive more support from the public. Furthermore, people's system support might directly affect government policy implementation, especially during the COVID‐19 pandemic, in terms of factors such as social distancing requirements, travel restrictions, and the isolation of infected individuals, which is crucial for the protection of public health. Thus, governments should try to administer responses that reflect control and adopt more effective response methods when facing a public emergency crisis.

This study has several limitations that can be addressed in future research. First, correlative studies do not indicate a causal link between governments’ responses and system support. Therefore, future studies could use an experimental method to provide confirmation of such a causal relationship. Second, the causes of discrepancy between the national and local systems were examined only from the perspective of governments’ responses in China. Future studies based on other countries from a cultural perspective and are necessary. Third, the whole system structure includes not only national and local political systems, but also others such as community and family, which are important for people to cope with the pandemic. Therefore, exploring the effects of different systems on individuals’ mindsets could provide meaningful insight.

OPEN RESEARCH BADGES

This article has earned Open Data and Open Materials badges. Data and materials are available at https://osf.io/qr4gp/?view_only=cd8c0960c94d444f9949b10345cfeedf.

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.