PARTICIPATION IN ANTI-JAPANESE DEMONSTRATIONS IN CHINA: EVIDENCE FROM A SURVEY ON THREE ELITE UNIVERSITIES IN BEIJING

Abstract

Nationwide anti-Japanese demonstrations have erupted in China periodically in recent years. This study investigates what factors make university students more motivated to participate in anti-Japanese demonstrations. We collected original data on 1,458 university students in Beijing in June 2014, inquiring about both actual and possible future participation. We find that students are more willing to participate in future demonstrations (1) when they believe that anti-Japanese demonstrations benefit China's diplomacy (instrumentality), and (2) when they have prior demonstrators in their social networks (diffusion). However, when it comes to actual participation, only diffusion plays a significant role while instrumentality does not. While students claim that they are motivated by beliefs that demonstrations will matter for China's diplomacy, they actually turn out only when networks operate. In addition, membership in the Chinese Communist Party (CCP) does not affect prospective participation but deters actual participation. The CCP actually discourages participation in recent anti-Japanese demonstrations.

INTRODUCTION

China and Japan are two important countries, but their relationship can be very turbulent. Anti-Japanese mass demonstrations have erupted in China periodically in recent years. The most recent wave of demonstrations occurred in August and September 2012, when anti-Japanese protests broke out in more than 100 cities across China. ¹ It was the largest wave of anti-Japanese mass demonstrations since China and Japan established diplomatic relations in 1972.

The scale and swift development of these demonstrations often shock outside observers, and have drawn much attention from international relations scholars and political scientists. Political scientists attribute modern-day anti-Japanese demonstrations to Japan's ambivalence toward its war responsibilities during World War II, insincere apologies for wartime atrocities, and conservative domestic politics (Benfell 2002; Berger 2003, 2008; Ienaga 1993; Kristof 1998; Lind 2008; Nozaki 2005; Orr 2001; Penney and Wakefield 2008; Wakamiya 1999), or to China's anti-Japanese patriotic education, surging nationalism, and the Chinese government's manipulation of anti-Japanese sentiments for instrumental purposes (Barme 1993; Brittingham 2007; Callahan 2006; Coble 2007; Cohen 2002; Gries 2004; He 2007; Mitter 2000, 2003; Reilly 2012; Shirk 2007; Wang 2008, 2012; Weiss 2014; Zhao 2005). By contrast, international relations scholars view the occurrence of anti-Japanese demonstrations in China as a result of the shifting international geopolitics and intensified conflict in national interests (Bush 2010; Calder 2006; Drifte 2002; Hughes 2009; Pyle 2007; Sasaki 2010; Söderberg 2002; Wan 2006).

Despite considerable interest from international relations and political science, there has been a surprising lack of sociological research on anti-Japanese demonstrations in China. As a result, the existing literature on China's anti-Japanese demonstrations is largely at the macro (state) level, with scarce attention paid to the individual level. In particular, a key sociological question has gone unanswered: what individuals are more likely to participate in anti-Japanese demonstrations? This study is intended to answer this question and investigate both possible future participation (intention) and actual participation (action). More importantly, it identifies and compares the motivating factors underlying prospective and actual participation.

In light of social movement theories, we highlight two motivating factors, instrumentality and diffusion. They reflect two distinct perspectives on the formation of individuals' motivation for movement participation. The instrumentality factor focuses on individuals' rational calculation and emphasizes the motivating force of achieving desirable movement goals (Huber 1997; Klandermans 1984; Marini 1992; Muller and Opp 1986; Oberschall 1994). In contrast, the diffusion factor is more socially oriented and stresses the influence of prior participants in one's interpersonal networks (Gould 1993; Knoke 1990; McAdam 1999; McAdam and Paulsen 1993; Snow, Zurcher, and Ekland-Olson 1980). In addition to these two perspectives, we also examine other possible factors such as organizational membership, perceived risks, nationalism, and socio-demographic characteristics.

It remains an open question whether these factors successfully explain individuals' prospective and actual participation in anti-Japanese demonstrations. Moreover, the factors shaping individuals' prospective participation may differ from those driving actual participation. Willingness to participate does not necessarily translate into actual participation, as collective action problems need to be solved. Some factors are more effective in overcoming collective action problems than others. The attitudes of the state may also matter for actual participation more than prospective participation. The Chinese Communist Party (CCP) can mobilize uninterested students to protest when it encourages such protests. It may also prevent motivated students from participating when it discourages protests.

To address these questions, we conducted a large-scale survey of 1,458 Chinese students from three top universities in Beijing, including Peking University (PKU), Tsinghua University (THU), and Renmin (People's) University of China (RUC), in June 2014. We chose university students as our subjects because they are “a highly active population and have played an important role in most major social movements” (Van Dyke 2003, 245). The role of university students in social movements is particularly prominent in the Chinese context. University students in Beijing have long played an essential role in all major social movements (Yu and Zhao 2006; Zhao 1998, 2001). The three universities were chosen because they are “among the largest and the most prestigious with a history of influencing other universities and Chinese politics in general” (Yu and Zhao 2006, 1758).

Analysis of this original data set reveals two key findings. First, while both instrumentality (perceived benefits of demonstrations for China's diplomacy) and diffusion (knowing prior demonstrators in one's social networks) both motivate prospective participation, diffusion is more effective in solving collective action problems than instrumentality. While students claim that they are motivated by beliefs that protests will matter for China's diplomacy, in fact they are more likely to turn out only when social networks operate. Second, CCP membership has no consistent effect on prospective participation but deters actual participation. The CCP has been inconsistent in its stance and directives on anti-Japanese demonstrations, but in recent years it actually discourages participation, at least in Beijing.

BACKGROUND ON ANTI-JAPANESE DEMONSTRATIONS

Relations between China and Japan have been tense in recent years due to historical and territorial disputes. Both countries often politicize historical memories about Japan's invasion of China during World War II. The relationship is further strained by the long-contested sovereignty over an island chain, the Diaoyu or Senkaku islands, in the waters between the two countries. China's dramatic social transition over the years adds more layers to the complexity of Sino-Japanese relations. Prior to the 1980s, the totalitarian government virtually excluded the general public from the making of policies related to Japan (He 2007, 2009; Rose 2005). As a result, popular concerns and grievances within the Chinese society were not well addressed in China's previous diplomacy with Japan. China's reforms since the late 1970s have eased political constraints on the society and facilitated much broader societal involvement in Sino-Japanese relations (Glaser and Saunders 2002; Gries 2004, 2005; Lampton 2001; Mertha 2009; Reilly 2012; Rose 2005; Tang 2005; Yang 2009; Zheng and Wu 2005).

One prominent manifestation of this increased societal involvement is the periodic occurrence of anti-Japanese demonstrations. Large-scale anti-Japanese demonstrations occurred in 1985, 2005, 2010, and 2012. ² The scale of these demonstrations has been gradually increasing over time. In August and September 2012, anti-Japanese demonstrations erupted throughout China in more than 100 cities, protesting the Japanese government's decision to purchase and “nationalize” three of the disputed Diaoyu or Senkaku Islands. Some observers claim that protesters took to the streets in as many as 208 of China's total 287 prefectural cities (Wallace and Weiss 2015; Weiss 2014). These street demonstrations were largely peaceful at the beginning, but in some places they turned violent. Japanese businesses were ransacked, windows broken, and cars smashed.

University students have been “a particularly common source” of participants and were “noticeably present” in recent anti-Japanese demonstrations (Wallace and Weiss 2015, 409). The scene of university students marching, holding banners and shouting slogans is a typical image of anti-Japanese demonstrations. Although there seemed to be a decline of student participation in the 2012 anti-Japanese demonstrations in first-tier cities such as Beijing and Shanghai, this is largely due to effective preventive measures taken by the government in these cities. Presence of a university student population is one of the most reliable predictors of the occurrence of anti-Japanese demonstrations (Wallace and Weiss 2015). It is thus important to investigate what factors motivate university students to participate.

Despite considerable interest in China's anti-Japanese demonstrations from both the academic community and the general public, there has been a surprising lack of research on the social bases of anti-Japanese demonstrations. We know little about which Chinese are more likely to participate. We have little knowledge about the composition of the “latitude of acceptance,” that is, the pool of potential recruits for the movement (Klandermans and Oegema 1987; McAdam and Paulsen 1993). Although not everyone in the pool eventually joins the demonstrations, individuals in this pool are more likely to get involved, and they constitute the fertile soil for future demonstrations. This study is the first empirical attempt to study which university students are more motivated to join anti-Japanese demonstrations and, more importantly, to reveal the factors driving both prospective and actual participation.

DETERMINANTS OF THE PARTICIPATION MOTIVATION

Individuals' motivation to participate and its determinants should be an essential part of any social movement studies (Dixon and Roscigno 2003; Walder 2009). Various factors affect which individuals are more likely to participate than others. From the scholarship on social movements, we identify several factors that may influence individuals' motivation. We pay special attention to two major perspectives, instrumentality and diffusion. They represent two competing perspectives on movement participation. In the instrumentality perspective individuals are motivated by their individual rational calculation, while in the diffusion perspective individuals are socially motivated by their interpersonal networks.

Instrumentality

The instrumentality perspective stresses rational perceptions of movement participation and suggests that individuals' motivation depends on perceived instrumental values of the movement (Huber 1997; Klandermans 1984; Marini 1992; Muller and Opp 1986; Oberschall 1994). In social movements that struggle for collective (rather than individual) goals, there should be more incentive to participate if the movement is seen as instrumental in improving the situation of the group.

Participation in anti-Japanese demonstrations may not bring tangible personal benefits, but individuals can be motivated in the name of “we” to pursue public goods (Muller and Opp 1986). Individuals may identify with the interests of the nation and consider how demonstrations would benefit the nation as a whole. In China, the state has the capacity to prevent, tolerate, or even promote anti-Japanese demonstrations. According to some scholars (Reilly 2012; Weiss 2014), the Chinese state is smart in that it manages the anti-Japanese movement for instrumental purposes. When the state seeks to gain leverage in diplomatic negations, it tolerates or even encourages anti-Japanese demonstrations. It can point to nationalist sentiments and popular protests and say that its hands are tied at the negotiation table. A typical example is the state's tolerance of anti-Japanese demonstrations in 2005 when thousands of Chinese took to the streets to oppose Japan's bid for a permanent seat on the United Nations(UN) Security Council. Those popular protests helped the Chinese state make a strong case against Japan's bid.

Reflecting this strategic interaction between the state and the society, Chinese citizens may consider the instrumental value of anti-Japanese demonstrations. We conceptualize instrumentality here as the foreign policy effect of the demonstrations. If students believe that their participation in anti-Japanese demonstrations would help the state's diplomatic policies, they should have more incentive to participate. Perceived instrumentality has long been found to be a good predictor of participation in collective action (Craig 1979; McAdam and Paulsen 1993; Sutherland 1981). Perceived instrumentality of anti-Japanese demonstrations should elevate individuals' motivation to participate.

Diffusion

The diffusion perspective on movement participation emphasizes how influences derived from interpersonal networks affect individuals' motivation to participate (Gould 1993; McAdam 1999; Snow, Zurcher, and Ekland-Olson 1980). This perspective locates individuals in their social networks and examines social influences from interpersonal ties.

Many scholars have found the importance of interpersonal ties in motivating individuals to participate in a movement (Hedström, Sandell, and Stern 2000; Knoke 1990; McAdam and Paulsen 1993; Snow, Zurcher, and Ekland-Olson 1980; Walgrave and Wouters 2014; Zhao 1998). Knowing someone already involved in a movement is one of the most reliable predictors of movement participation. Mobilization is seen as essentially a diffusion process primarily occurring in social networks. Interpersonal ties influence a person's participation motivation through two intertwined mechanisms: imitation and persuasion (Hedström, Sandell, and Stern 2000). First, individuals' observation of others' behavior in their networks often leads to imitation, as individuals dislike looking different from their peers and have a tendency to learn from the experiences of others they know. They often imitate others in their networks consciously or subconsciously so their subsequent behavior becomes similar to that of others. Second, the presence of a tie to someone already involved in a movement helps individuals receive more information about the movement. They are also more likely to encounter social interactions that persuade or invite them to participate. Taken together, interpersonal ties serve as a conduit for imitation and persuasion that elevate people's motivation to join demonstrations. Those who know someone with experience in anti-Japanese demonstrations are more likely to participate in anti-Japanese demonstrations.

Organizational Membership

Scholars find membership in formal organizations (movement-related or not) increases individuals' motivation to participate in social movements through two intertwined mechanisms. First, experience with organizations is presumed to forge social ties within organizations that may facilitate people's motivation in social movements (McAdam 1986; McAdam and Paulsen 1993; Schussman and Soule 2005). Second, engagement in organizations cultivates civic skills that make individuals feel more ready to participate (Brady, Verba, and Schlozman 1995; Verba, Schlozman, and Brady 1995). These civic skills include organizational and communication skills and enhanced political interests. More civic skills should increase people's motivation for political participation of any kind including demonstrations (Verba, Schlozman, and Brady 1995). In the Chinese context, the most common forms of formal organizations that university students can engage in are student organizations and the CCP (Yu and Zhao 2006). Those who are members of student organizations and the CCP may be more motivated to participate in anti-Japanese demonstrations.

Although the literature points to a positive effect of organizational membership on student participation, it remains an empirical question whether these organizations actually promote movement participation in the Chinese context. The effect of organizational membership is likely to differ for prospective and actual participation. Its effect on actual participation may be particularly dependent on the stance of the state. Social organizations in China are heavily controlled and monitored by the state. Membership in these organizations, CCP membership in particular, can make it easier for the state to mobilize students to protest (when it encourages such protests) or prevent students from participating (when it discourages protests). ³ Hence, in some cases membership in these organizations may deter participation, especially actual participation, when the state feels threatened by such collective actions.

Other Factors: Perceived Risks, Nationalism, and Normative Justification

We also consider and control for other relevant rational and political factors such as perceived risks, collective identity and normative values. In addition to instrumentality, another indispensable component in rational calculation is perceived risks that deter individuals' motivations to join a social movement. Anti-Japanese demonstrations can potentially trigger domestic instability and pose a threat to the Chinese state itself. Given the limited channels for political participation in China, citizens may seize the opportunity to promote their domestic goals and vent anger that goes beyond anti-Japanese sentiments. It has been observed that anti-Japanese protesters sometimes turn their anger to the Chinese state and advance domestic objectives (Reilly 2012; Wallace and Weiss 2015; Weiss 2014). Maintaining domestic stability (Weiwen) has become a top political priority for the Chinese state. When the perceived risk of jeopardizing domestic stability is high, the state is more likely to prevent and suppress anti-Japanese demonstrations. If individuals believe that anti-Japanese demonstrations would jeopardize China's domestic stability, they should be less willing to participate due to perceived high risks stemming from instability. Domestic instability will not only disrupt social order but also potentially invite state repression. Potential state repression would make participation riskier and increase the cost of participation, thereby decreasing individuals' motivation to participate.

Collective identity and normative values both motivate individuals in social movements (Benford and Snow 2000; Klandermans 1984; Polletta and Jasper 2001; Snow and Benford 1992). A movement is more appealing if it is congruent with a person's cultural values, such as collective identity and moral standards. In particular, nationalism often provides a potent form of collective identity and looms large in movements against foreign targets (Calhoun 1993; Kane and Park 2009). Nationalist identity is particularly salient in authoritarian and transitional countries (such as China) “where diverse institutions of civil society are lacking or fail to provide for a diversity of public discourses and collective identities” (Calhoun 1993, 387). Indeed, nationalism is often considered as an important source of anti-Japanese sentiments in China (Gries 2004, 2005; Reilly 2012; Wang 2008; Zhao 2004b, 2013). Individuals with strong nationalist identity proudly identify themselves as part of the nation and develop emotional attachments to the nation (Goodwin, Jasper, and Polletta 2001; Polletta and Jasper 2001). Anti-Japanese demonstrations resonate with this nationalist identity. They may feel more obligated and more emotionally motivated to demonstrate against Japan to defend perceived national interests.

Normative perceptions of an issue at stake also matter for individuals' motivation to join the movement for this issue. The same issue, especially its cause, is evaluated differently by different people. The motivation to participate can be enhanced by normative perceptions concerning the causes of the discontent (McAdam 1999; McAdam and Paulsen 1993). Individuals are more likely to be motivated if they perceive the situation as unjust and their protested target as the reason for this injustice. If individuals see themselves as occupying the moral high ground and put the blame on others, they would likely possess a sense of righteousness that can transform into righteous anger against others (Stürmer and Simon 2009) and provide moral justification for protesting behavior (Kane and Park 2009). In the case of Sino-Japanese relations, the two countries often engage in a blame game and accuse each other of wrongdoing in historical and territorial disputes. If individuals consider Japan as culpable for the troubled Sino-Japanese relations, they should feel more morally justified to protest.

Socio-Demographic Characteristics

Some basic socio-demographic characteristics may shape students' motivation to demonstrate in the Chinese context (Walder 2006; Yu and Zhao 2006; Zhao 2003). Gender and family income are often used in predicting students' movement participation. It is often found that due to gendered socialization, men on average are more willing to take risks, whereas women are relatively more averse to risks. As all street demonstrations entail some degree of risk, men are more likely to protest than women (Caren, Ghoshal, and Ribas 2011; Dalton 2006). This gender difference may be true in China's anti-Japanese demonstrations too.

Protests are commonly rooted in social inequalities. Differential participation often reflects different attitudes between those with vested interests in the status quo and those who seek to change it (Dixon and Roscigno 2003; Walder 2006). Individuals from less well-off family background have less interest in maintaining the status quo and are more willing to participate in radical social movements that call for change. We expect that students with lower family income are more likely to participate in anti-Japanese demonstrations.

Although we do not have definite expectations about their effects, we also consider other socio-demographic characteristics, such as ethnicity (Han versus non-Han), household registration (Hukou) type (rural versus urban), and hometown (the region where they live before coming to Beijing). There may be differences between the Han Chinese and other minority ethnical groups in their participation motivation. Given the prominent divide between rural and urban residents and vast regional diversity in China, it is interesting to examine the effects of household registration types and regional differences too. It remains an open empirical question whether the motivation to participate in anti-Japanese demonstrations varies along the lines of ethnicities, household registration types, and regions.

DATA AND METHOD

To examine the factors underlying university students' motivation to participate in anti-Japanese demonstrations, we conducted a large-scale survey of 1,458 students from three top universities in Beijing, including Peking University (PKU), Tsinghua University (THU) and Renmin (People's) University of China (RUC), in June 2014. We collected information on students' prospective and actual participation in anti-Japanese demonstrations and the factors that may potentially affect participation.

We conducted the survey on the three universities because of their prominent role in social movements in the modern Chinese history. In almost all major social movements since the 1900s, university students in Beijing have been a pioneering force (Zhao 1998, 2000, 2004a). Especially, PKU, THU, and RUC have been recognized as the center of student activism in China (Yu and Zhao 2006; Zhao 1998, 2004a). In 1919 thousands of PKU students launched anti-Japanese demonstrations against the Treaty of Versailles that forced China to grant territories and special privileges to Japan. These demonstrations triggered the May Fourth Movement that influenced the course of China's modern history. Facing Japan's military aggression, thousands of students in Beijing again marched on December 9, 1935, which turned into a nationwide movement resisting the Japanese invasion (Zhao 2000). In 1989, students from many Beijing universities protested the Communist government and demanded political reform. Students of PKU, THU, and RUC played a central role in student mobilization (Deng 1997; Wright 2001; Zhao 2004a). Similar prominent roles of Beijing university students, especially from these three universities, can also be seen in the anti-US demonstrations in 1999 (Yu and Zhao 2006) and the anti-Japanese demonstrations in 1985, 2005, and 2010 (Reilly 2012; Rose 2005; Weiss 2014). ⁴ In light of their significance in China's social movement history, we selected these three universities for our survey.

Sample and Sampling

We selected our survey participants according to carefully designed probability sampling. We recruited a survey team from each university. Each team consisted of 5–7 interviewers and surveyed about 600–700 students from their respective university. Finally, a total of 1,458 questionnaires were returned (504 from PKU, 467 from THU, and 487 from RUC). The overall response rate is 72.5 percent. Our sample includes students from all the 22 provinces, 5 autonomous regions, 4 municipalities, 2 special administrative regions (Hong Kong and Macau), and Taiwan. In the sample, the average age is 22 years. 54.05 percent of participants are male (N = 788) and 45.95 percent are female (N = 670). Undergraduate and graduate students constitute 67 percent and 33 percent of the participants, respectively. The participants in the sample are predominantly Han nationals (89.71 percent), while minority nationals make up the remaining 10.29 percent. These numbers well reflect the overall profile of the student body in these universities. ⁵ Careful sampling gives us confidence in the sample's representativeness of the student population at the three universities.

Variables

Dependent variables

We use two dependent variables that tap into students' prospective and actual participation in anti-Japanese demonstrations, respectively. One dependent variable is students' willingness to participate in future anti-Japanese demonstrations, or prospective participation. We use this survey question to measure this variable: “If there are demonstrations against Japan in the future, how likely will you participate?” The possible responses are on a 5-point scale ranging from “very unlikely” (1) to “very likely” (5). A higher score indicates a higher level of the motivation to participate in future anti-Japanese demonstrations. Figure 1 depicts the distribution of the motivation among students. Students' motivation varies along the 5-point scale. A substantial proportion, namely 45.64 percent (6.36% + 39.28%), of all sampled students express definite willingness to participate in future anti-Japanese demonstrations.

Figure 1 Participation in Anti-Japanese Demonstrations

The second dependent variable is students' actual experience in anti-Japanese demonstrations, or actual participation. We use this question to capture this variable: “Have you ever taken part in any demonstrations against Japan?” It is a binary variable with 1 indicating “yes” and 0 “no.” As shown in Figure 1, only 63 out of the 1,458 students surveyed (or 4.32%) have actually participated in anti-Japanese demonstrations.

Instrumentality

We measure instrumentality through the question: “Do you think anti-Japanese demonstrations are beneficial to China's diplomacy toward Japan?” The respondents have five possible responses on a 5-point scale ranging from “not beneficial at all” to “highly beneficial.” A higher score suggests a higher level of perceived instrumental values of anti-Japanese demonstrations.

Interpersonal ties

This is a binary variable. Respondents are asked whether they know someone who has experience of participating in anti-Japanese demonstrations. If they know, it is coded as 1. If not, it is coded as 0.

Membership in student organizations

We use the frequency of participation in student organizations. It is measured on a 3-point scale including “no participation” (1), “occasional participation” (2), and “regular participation” (3). A higher score indicates a higher level of involvement in student organizations.

Membership in the CCP

CCP membership is a binary variable, with 1 indicating a CCP member and 0 a non-CCP member.

Perceived risks

Perceived risk is measured by this question: “Do you think anti-Japanese demonstrations would jeopardize China's domestic stability?” Responses are coded on a 3-point scale including “promoting domestic stability” (1), “no effect on domestic stability” (2), and “jeopardizing domestic stability” (3). A greater number indicates a higher level of the perceived negative impact of anti-Japanese demonstrations on domestic stability, which implies greater concern about the risk of participating in demonstrations.

Nationalist identity

We include two questions measuring nationalist identity. The first question is “To what extent do you agree or disagree with the following statement that I am proud to be a Chinese rather than a citizen of other countries.” The second one is “To what extent do you agree or disagree with the following statement that the world would be better if all other countries were like China.” We code people's answers on a 5-point scale from “disagree strongly” (1) to “agree strongly” (5). Therefore, for both questions a higher score implies stronger nationalist identity. We use the variable from the first question in the following analysis. However, we also tried using the variable from the second question and the results were substantively the same.

Normative justification

Normative justification is captured by this question: “Which side do you think should be responsible for the troubled Sino-Japanese relations?” We measure the responses on the following 5-point scale: “China should take full responsibility” (1), “China takes primary responsibility while Japan takes secondary responsibility” (2), “China and Japan take equal responsibility” (3), “Japan takes primary responsibility while China takes secondary responsibility” (4), and “Japan should take full responsibility” (5). Hence, a higher score suggests a greater level of blame respondents put on Japan for Sino-Japanese disputes.

Other socio-demographic characteristics

Gender is a binary variable, with male coded as 1 and female as 0. Family income is measured through this question: “How much is your parents' total monthly income approximately?” The possible responses are on a 9-point scale—lower than 1,000 RMB ⁶ (1), 1,000–2,000 RMB (2), 2,000–4,000 RMB (3), 4,000–6,000 RMB (4), 6,000–8,000 RMB (5), 8,000–10,000 RMB (6), 10,000–15,000 RMB (7), 15,000–20,000 RMB (8), and over 20,000 RMB (9). A larger number indicates more family income. Ethnicity is measured as a binary variable, with the Han ethnic group coded as 1 and the non-Han group as 0. Household registration or Hukou type is also binary, with rural residents coded as 1 and urban residents as 0. Region is measured by a series of dummy variables including East China, Northeast China, North China, Central China, South China, Northwest China, Southwest China, and other (areas outside mainland China, including Hong Kong, Macau, and Taiwan). ⁷

Table 1 shows basic descriptive statistics of the variables. In terms of the socio-demographic distribution, in our sample (1) 54.0 percent of all respondents are male while 46.0 percent are female; (2) 31.4 percent are CCP members while 68.6 percent are not; (3) 89.7 percent are of Han ethnicity while 10.3 percent belong to non-Han minority groups; (4) 21.8 percent are of rural household registration type while 78.2 percent are urban residents; (5) the percentages of all respondents coming from East China, Northeast China, North China, Central China, South China, Southwest China, Northwest China, and other (including Hong Kong, Macau, and Taiwan) are 27.0 percent, 9.7 percent, 19.2 percent, 16.1 percent, 7.9 percent, 11.4 percent, 7.7 percent, and 1.0 percent, respectively; (6) the average monthly family income of all respondents is approximately between 6,000–8,000 RMB.

Table 1 Descriptive Statistics for Variables Used in the Analysis

Method

We use the OLS regression to analyze prospective participation. We define Y as the level of prospective participation in anti-Japanese demonstrations, and specify the model as follows:

Y = β₀ + β₁Instrumentality + β₂Tie + β₃Membership + β₄CCP + β₅Nationalist + β₆Justification + β₇Risk + β₈Sex + β₉Income + β₁₀Ethnicity + β₁₁Household + β₁₂Region + ε

where β₁-β₁₂ are the coefficients of the explanatory variables and ε is the error term.

Actual participation is a binary dependent variable and is thus analyzed by the logistic regression. We define p as the probability of the binary dependent variable equal to 1 (in this case, the probability that the respondent has actually participated in anti-Japanese demonstrations) and let p be modeled using a logit link function. The model is specified as follows:

Log[P/(1 − P)] = β₀ + β₁Instrumentality + β₂Tie + β₃Membership + β₄CCP + β₅Nationalist + β₆Justification + β₇Risk + β₈Sex + β₉Income + β₁₀Ethnicity + β₁₁Household + β₁₂Region + ε

where β₁-β₁₂ are the coefficients of the explanatory variables and ε is the error term.

We use the Stata software (release 12) (StataCorp 2011) in the modeling. We employ the AIC and BIC measures to identify models with good fit. They are popular indicators for model selection (Akaike 1974; Raftery 1995). They assess the goodness of model fit while penalizing the number of variables included. When several models are estimated on the same data, the one with smaller AIC and BIC values is considered to be better. For each model we calculate these model fit statistics and report them in the tables. In the regression we also test multicollinearity and do not identify problems. ⁸

RESULTS

Prospective Participation

We first estimate a series of OLS regression models to investigate what factors make university students more willing to participate in future anti-Japanese demonstrations. Results are presented in Table 2. We start with a simple model (Model 1) that contains basic socio-demographic variables only. Family income and coming from outside mainland China have significant and negative effects on prospective participation, whereas gender, ethnicity, and household type do not show any significant effects. Students from less well-off family background are more willing to join future anti-Japanese demonstrations. In contrast, there are no significant differences between male and female students, between Han Chinese and non-Han students, or between rural and urban students. There are no significant differences among students from different regions within mainland China, but students from outside mainland China such as Hong Kong, Macau, and Taiwan do significantly show a lower level of motivation.

Table 2 OLS Regression of Prospective Participation in Anti-Japanese Demonstrations

Note: (1) N = 1458; (2) Numbers in parentheses are standard errors; (3) From 2-tailed tests, * P < .05; ** P < .01; *** P < .001; (4) For “region” dummy variables, in Models 1 and 2 East China is the reference group and thus omitted, while in Models 3 and 4 the reference group becomes all regions within mainland China.

These basic socio-demographic variables combined can only account for 3.2 percent of the total variation among students in their prospective participation, according to the coefficient of determination (R²) for Model 1. Next we bring in all other explanatory variables and estimate Model 2. Consistent with Model 1, among basic socio-demographic variables only family income and coming from outside mainland China have significant effects. Here the effects of other newly added variables are of major interest. In comparison with Model 1, including these additional variables does greatly decrease the AIC and BIC values, suggesting a substantial increase in model fit. Now as much as 24.5 percent of the total variation in students' prospective participation is explained by Model 2. Therefore, including these additional variables does greatly increase the capacity of the model in explaining differential motivation among students.

First, instrumentality has a significantly positive effect on prospective participation in anti-Japanese demonstrations. Students who see anti-Japanese demonstrations as benefiting China's foreign policies are more motivated to participate. Conversely, perceived risks have a significantly negative effect on prospective participation. The perception of anti-Japanese demonstrations posing a threat to China's domestic stability deters the motivation for movement participation.

Second, interpersonal ties display a significant and positive effect on prospective participation. Knowing someone who has participated in anti-Japanese demonstrations significantly elevates a student's motivation to participate.

Third, memberships in both student organizations and the CCP have no significant effect on prospective participation. Both variables show positive effects, but they fail to reach statistical significance. Involvement in formal organizations makes no significant difference in students' motivation. The lack of a consistent effect of organizational membership may reflect variation in CCP directives, sometimes discouraging, sometimes allowing or even encouraging participation.

In addition, both nationalist identity and normative justification display significantly positive effects. Students who have a higher level of pride in being Chinese are more likely to join future anti-Japanese demonstrations. Those who blame Japan for the disputes between China and Japan are also more motivated.

We further apply the stepwise (forward) modeling procedure to the modeling. Instead of specifying a model with certain variables ex ante, we model the process and let the stepwise model selection technique to select significant variables into the final model. This resultant model confirms the findings from previous models. Without those insignificant variables in the modeling, the R² only decreases slightly from .245 (Model 2) to .240 (Model 3), indicating that the selected variables in Model 3 can collectively explain 24 percent of the total variation among students. Model 3 sacrifices little explanatory power while including a much more parsimonious set of explanatory variables in the modeling. In comparison with previous models, Model 3 generates the smallest AIC and BIC values, suggesting the best model fit.

To compare the relative importance of these significant variables, we further estimate a model that uses standardized coefficients. We first rescale all explanatory variables and transform them into standardized scores (with a mean of zero and a variance of one). Then we re-estimate Model 3 with these standardized variables. Results are presented in Model 4. Standardized variables are unit-free. By comparing standardized coefficients we can examine which variables have greater effects. Among the significant variables, the effects of instrumentality and interpersonal ties stand out, outranking all other variables. Instrumentality and interpersonal ties are indeed two most influential factors motivating students to join future demonstrations.

Actual Participation

We also estimate a battery of logistic regression models to examine the factors driving students' actual participation in anti-Japanese demonstrations. Results are presented in Table 3. The modeling strategy is similar to that used for prospective participation.

Table 3 Logistic Regression of Actual Participation in Anti-Japanese Demonstrations

Note: (1) N = 1458; (2) Numbers in parentheses are standard errors; (3) From 2-tailed tests, * P < .05; ** P < .01; *** P < .001; (4) For “region” dummy variables, East China is the reference group and thus omitted.

Model 1 contains basic socio-demographic variables only. None of these variables have significant effects, except two region dummy variables, North and Southwest China. In comparison with East China, students from North China and Southwest China are more likely to have actually participated in anti-Japanese demonstrations. Nevertheless, when other variables are added into the following models, the significant effects of these region dummy variables disappear. Overall, basic socio-demographic variables cannot well explain students' actual participation. According to the R², these variables combined can only explain 2.9 percent of the total variation in students' actual participation.

Model 2 includes all explanatory variables and improves greatly over Model 1 in its capacity of explaining actual participation. Now 24.7 percent of the variation in actual participation can be explained. Among all explanatory variables, only three show significant effects. First, interpersonal ties have a significantly positive effect. Knowing someone who has participated in anti-Japanese demonstrations increases the likelihood of a student actually participating too. Second, CCP membership shows a significant and negative effect, implying that CCP members are less likely to have actually participated in anti-Japanese demonstrations. In addition, nationalist identity also displays a significantly positive effect, suggesting that a higher level of nationalism makes a student more likely to actually participate. It is worth noting that instrumentality has no effect on actual participation.

Model 3 presents the result from the stepwise (forward) modeling procedure that selects significant variables. The finding is consistent with Model 2—only interpersonal ties, CCP membership, and nationalism have significant effects and are thus selected into the final model. The three variables alone can explain as much as 22.6 percent of the variation in actual participation. The AIC and BIC values are much lower than those of the previous models and thus indicate an improved model fit. We further present standardized coefficients in Model 4. Interpersonal ties have the strongest effect, followed by CCP membership and nationalism.

DISCUSSION AND CONCLUSION

Anti-Japanese demonstrations have already become one of the largest social movements in today's China. Nevertheless, unlike political scientists and international relations scholars, sociologists have been slow in responding to the great interest in these demonstrations from both the academic community and the general public. This study is the first empirical investigation of the factors shaping individuals' participation in anti-Japanese demonstrations. The analysis of the original data collected from three elite universities in Beijing generates important insight. We distinguish prospective and actual participation, and find interesting divergence in their motivating factors.

Prospective Participation

As for prospective participation (willingness to participate in future anti-Japanese demonstrations), the most influential motivating factors are instrumentality and diffusion. If students perceive anti-Japanese demonstrations as instrumental in benefiting China's Japan policies, or if they have ties with prior demonstrators, they are more willing to participate in future anti-Japanese demonstrations. In contrast, basic socio-demographic characteristics such as gender, ethnicity, and household registration (Hukou) type show no significant effects. There are no regional differences within mainland China as well. Membership in formal organizations such as student organizations and the CCP does not significantly increase students' motivation either. This insignificant effect of membership may reflect inconsistency in the CCP's stance and directives on anti-Japanese demonstrations—sometimes discouraging or suppressing, sometimes allowing or even encouraging participation. Nevertheless, family income has a significant effect, and the difference between mainland and non-mainland students is also significant. Students from less well-off families are more motivated to join future anti-Japanese demonstrations. Students from mainland China also show a higher level of motivation than those from outside mainland China. In addition, we also find that strong nationalist identity and blaming Japan for troubles between China and Japan significantly elevate the motivation, while worrying about undesirable consequences such as domestic instability deters the motivation.

Actual Participation

Only three factors, interpersonal ties, CCP membership, and nationalist identity, are found to be significant in students' actual participation in anti-Japanese demonstrations. Not surprisingly, nationalism plays a significant role in driving actual participation. Social networks also have a significantly positive effect, indicating that social networks mobilize students to participate. CCP membership has a significant and negative effect on actual participation, so CCP members are less likely to actually participate in comparison with non-CCP members. This is consistent with the anecdotal observation (Wallace and Weiss 2015) that in recent anti-Japanese demonstrations the government took some measures that effectively prevented students from staging large-scale anti-Japanese demonstrations in Beijing. Since the state has more effective control over its party members, it is no surprise that CCP members are less likely to have actually demonstrated. The negative effect of CCP membership implies the state's unfavorable attitudes toward recent anti-Japanese demonstrations. Instead of tolerating or promoting anti-Japanese demonstrations, the state actually discouraged or even suppressed participation in recent waves of demonstrations, at least in such top-tier cities as Beijing.

Prospective Participation versus Actual Participation

When comparing significant motivating factors underlying prospective and actual participation, we can see some commonalities. For instance, in both cases nationalist identity plays a significant role, indicating that nationalism looms large in both prospective and actual participation. The divergence between the determinants of prospective and actual participation is even more notable, however. While instrumentality and diffusion both promote prospective participation, only diffusion (network ties) matters for actual participation. Rational calculation of the foreign policy effect of protests may motivate students' expressed willingness to participate, but this willingness does not translate into action. From willingness to action students have to overcome the classic “free rider” obstacle in collective action. While some students realize that anti-Japanese demonstrations benefit China's diplomacy against Japan and are thus willing to participate, they may not actually participate. Typical thinking goes along the lines of “as only one individual it would not make a difference whether I participate or not.” Participating as a lone individual in street demonstrations can be lonely, boring, and even intimidating. Hence, the instrumentality motivator does not necessarily facilitate actual participation. In contrast, social networks not only elevate students' willingness to participate but also provide a powerful mobilizing tool. Individuals may feel more comfortable, empowered, or even compelled to participate in actual anti-Japanese demonstrations, when they are aware that others they know will participate too. Students are more likely to actually join anti-Japanese demonstrations in social groups rather than as lone rational individuals.

Another notable divergence is found in the effect of CCP membership. Whereas there is no difference between CCP and non-CCP members in their willingness to demonstrate against Japan, in practice CCP members are less likely to actually participate in anti-Japanese demonstrations. This divergence implies that political opportunities such as the state's attitudes are more important for actual participation than prospective participation. While many students are willing to protest, the state's attitudes (promotion, tolerance, or suppression) partly determine whether they have the political opportunity to turn this willingness into real action. Judging from the diverging effect of CCP membership, we contend that the Chinese state discourages or even suppresses students' participation in recent anti-Japanese demonstrations, at least in Beijing. The state has more effective control over CCP members through various party organs, and its preventive measures apply more easily to CCP members. CCP members face more pressure and obligations to adhere to state policies. Consequently, CCP members have to participate less than their non-CCP counterparts, although they do not differ in their willingness to participate.

Summary

These findings shed light on the profile of potential and actual student participants in anti-Japanese demonstrations from the three elite universities in Beijing. These potential demonstrators are nationalistic, as they expressively display strong nationalist identity and righteous anger. They are highly social in that they are subject to influences from prior demonstrators in their social networks. They are also rational. They strategically calculate the instrumental value of the demonstrations in promoting China's diplomacy. Nevertheless, when it comes to actual participation, rational calculation of instrumental values does not matter. The instrumentality-driven motivation does not translate into actual action. In contrast, nationalism and particularly social networks can actually mobilize students into collective action. Political opportunities allowed by the state are also important for translating intention into action. Taken together, these findings help us better comprehend the multifaceted nature of China's anti-Japanese demonstrations. While potential participants appear to be rational, nationalistic and social, actual anti-Japanese demonstrations should be understood as a mixed product of nationalistic fervor, social diffusion, and political opportunities.

Last but not least, we acknowledge and reflect on the limitations of this study. The survey was conducted in three elite universities in Beijing so the group under study is the student body from these universities. We should avoid overgeneralizing the findings here to all university students around China, let alone other social groups such as migrant workers in cities, white collar employees, high school students, and veterans, who also often participate in anti-Japanese demonstrations. Although we conjecture that many of the findings are not limited to the three elite universities, we leave the generalizability of our findings as an empirical question that requires further exploration. In this sense, this study also serves as a call for more empirical research on China's anti-Japanese demonstrations that applies our findings to a broader student population and to other social groups.