A spatial valence model of political participation in China

Abstract

In spatial models of political competition in democracies, citizens vote for the party or candidate that is the closest to their own ideological position, while in valence models, voters decide on the basis of non-policy factors, such as competence. What remains unclear, however, is whether citizens in authoritarian regimes use spatial or valence considerations to guide their decisions to participate in politics. This study uses data from the 2015 Chinese Urban Governance Survey to measure the ideology of Chinese citizens, and estimates an empirical stochastic model to explore how Chinese citizens use ideological distance and valence to determine how they want to participate in politics. The results show that valence issues, such as perceived government competence, play a larger role in political participation than ideology.

Keywords

China ideology political participation valence

1. Introduction

How do people make political choices under authoritarian rule? Spatial theories of political behavior in democracies model voting using the ideological distance between individuals and political parties, while valence theories argue that non-policy evaluations of a political actor are also important. Whether citizens in autocracies primarily participate in politics because of ideological or valence motivations is an open question. In this paper, I use a survey of Chinese citizens and an empirical stochastic model that incorporates both spatial distance and valence to venture an answer to this question. I find that while the spatial model does explain a significant part of the decision to consider protesting or the decision to join the Communist Party, valence, and in particular evaluations of the government’s competence, is a more important factor for explaining political participation in China.

The logic of the spatial model also predicts that political actors should locate themselves at a particular point in the ideological space to maximize their popular support. In the classic spatial model, which considers ideology along one dimension and models individual choice in a deterministic fashion, that point is the median voter (Downs, 1957; Riker and Ordeshook, 1973). In stochastic spatial models, parties are expected to converge on the electoral mean (McKelvey and Patty, 2006). These results are at odds with cases such as the USA, where political parties fail to converge on the median voter. To reconcile this divide between theory and outcomes, Schofield (2007) incorporates asymmetries in valence into the model. In the Schofield model, political parties do not necessarily converge on the electoral mean in equilibrium. Lower-valence political actors may be forced to move to the fringe of the ideological space to maximize their support.

Although the spatial model was designed to explain electoral politics, a similar spatial logic guides political contestation in authoritarian regimes. High-valence political actors, such as the government, attempt to occupy the center of the ideological space and paint the potential opposition as the ideological fringe (Schofield and Levinson, 2008). In my analysis, I find that ideological distances between Communist Party members and members of the potential opposition are relatively small, and that both groups locate themselves close to the center of the ideological space. This result suggests that citizens choose between the Communist Party and the potential opposition on the basis of valence issues, rather than ideological ones.

In the next section of this paper, I review previous research on the spatial model and generate theoretical expectations for how ideology and valence operate in China. Then, after sketching Schofield’s spatial valence model, I explain how I constructed my measures of ideology, estimate the model, and present the empirical results. The conclusion explores some of the implications of the findings for political contestation in China in the future.

2. Spatial and valence explanations of political behavior

In the classic spatial model popularized by Downs (1957) and Hotelling (1929), political parties are motivated by holding office and choose a policy position in the ideological space to maximize their share of the vote. Citizens vote for the party that has the policy position that is closest to their own views. Under this framework, political parties converge on the median voter, which leaves voters indifferent between their electoral choices.

One of the early objections to this line of reasoning was that not all issues lent themselves to variation along an ideological space. While public opinion is divided in its support for some policies, such as the proper level of state involvement in the economy, for certain valence issues, such as the need for honest leaders, or the need for competent administration, there is broad consensus among the public. When politicians campaign on valence issues, instead of taking specific policy positions, they attempt to associate themselves with some sort of positive symbol or goal, such as honesty or competence (Stokes, 1963, 1992). If they succeed in drawing some type of valence distinction between themselves and their opponents, then they may not need to converge to the same ideological position.

In recent years, scholars of mass–elite linkages have sought to formally combine spatial models of political competition with valence issues (Ansolabehere and Snyder, 2000; Groseclose, 2001). Empirical work in this literature has modeled vote choice probabilistically, using a mixed logit statistical model (Adams and Merrill, 1999; Adams et al., 2005; Micozzi and Saiegh, 2015; Schofield and Sened, 2005). By combining spatial and valence considerations, these models can help explain why parties fail to converge on the mean voter in some cases. In majoritarian or winner-take-all electoral systems, these models find that centripetal electoral forces tend to encourage political parties to converge on the center (Schofield et al., 2011a,b), but in proportional electoral systems, these models expect parties to diverge in equilibrium (Kurella and Pappi, 2015; Schofield and Sened, 2005; Schofield et al., 2011c).

Studies of hybrid regimes or electoral autocracies in this framework find that valence is an especially important factor in political behavior. In Russia, for instance, while ideological distance was a significant factor in the 2007 Duma elections, a voter’s opinion of Vladimir Putin was the most important factor for vote choice (Schofield and Zakharov, 2009). Similarly, in Singapore, ideological alignments are a lesser factor than valence issues, such as perceptions of party credibility, in voting decisions (Oliver and Ostwald, 2018). Valence differences also explain why political parties in electoral autocracies fail to converge on the mean voter. One common result is that lower-valence opposition parties are forced to the fringe of the ideological space, while the higher-valence ruling party occupies the center of the distribution (Schofield et al., 2011c, 2012). Because the government maintains control over the media in these states, opposition parties are often forced to use protests to express their discontent with government policy. This tends to make it difficult for the opposition to raise its valence in the eyes of the general population.

The empirical study of ideology under authoritarian rule is still in its early stages.¹ What remains unclear is whether a spatial logic guides the political behavior of citizens in personalist, military, or single-party authoritarian regimes. If citizens decide to support the regime or rebel because of the spatial distance between them and the government, then autocrats may feel the constraints of the public’s policy preferences, even in the absence of free and fair elections. However, if valence issues predominate, then an autocrat may be able to select an ideal point that is far from the average citizen and stay in power, so long as the regime maintains a valence advantage over the potential opposition.

Schofield and Levinson (2008) argue that political contestation in authoritarian regimes can be understood according to the logic of the spatial valence model. The autocrat generally attempts to capitalize on his high valence by occupying the center of the ideological space. In this way, he has the option of co-opting potential opposition by offering policy compromises. Dictators often lose power when they lose their valence advantage or allow the potential opposition to occupy the center of the ideological space.

3. Theoretical expectations

The leaders of the Chinese Communist Party (CCP) have often argued that incorrect ideological positioning undermines its control over society. One theme of official doctrine is the need to avoid veering too far to the ‘Left’ or the ‘Right.’

In a 1955 speech at a national Communist Party conference, Mao explained that: ‘to move far ahead of the times, to outpace current developments, to be rash in action and in matters of principle and policy and to hit out indiscriminately in struggles and controversies—these are “Left” deviations and are no good. To fall behind the times, to fail to keep pace with current developments and to be lacking in militancy—these are Right deviations and are no good either’ (Mao, 1977: p. 167).

As Mao grew frustrated with the bureaucracy’s resistance to his efforts to transform Chinese society, more and more officials began to find themselves accused of ideological deviancy. During the purges of the Anti-Rightist Movement and the Cultural Revolution, the range of acceptable ideological views converged to Mao’s purported positions (MacFarquhar and Schoenhals, 2006; Nathan and Shi, 1996).

After Mao’s death, the Party concluded that the Cultural Revolution had been a grave ‘Left’ error, which was both the product of Mao’s mistakes and an aberration inconsistent with Mao Zedong Thought (CCP Central Committee, 1981).² After ascending to power, Deng Xiaoping cautioned that the Party needed to return to the center. In his speech ‘Uphold the Four Cardinal Principles,’ he argued that ‘both the ultra-Left and Right currents of thought run counter to Marxism–Leninism and Mao Zedong Thought and obstruct our advance toward modernization’ (Deng, 1984: p. 173).

This rhetorical tradition leads us to a theoretical expectation for the Communist Party, which is that it will attempt to locate itself in the center of the ideological spectrum. If this is the case, then we might also expect that the individuals who choose to join the Party are relatively centrist in their ideological views, and that they become Party members for valence reasons. This expectation is in keeping with the Party’s strategy of preferentially enrolling the elite segments of Chinese society (Dickson and Rublee, 2000), and with the fact that many Party members join the Party to advance their career prospects and enjoy access to particularistic benefits (Dickson, 2014).

The motivations of the potential opposition are necessarily more opaque. Participation in collective action may be driven by ideological factors, such as opposition to the political system, or by valence issues, such as dissatisfaction with corrupt or inept administration.

Citizens who contemplate joining a protest must, moreover, grapple with the Communist Party’s efforts to design and restrict the scope of contention. In recent years, the Party has established a pattern of compensating some of the protesters who limit their demands to the resolution of localized grievances, while cracking down on those who put forward a broader agenda. By organizing its repressive strategies in this way, the CCP has preserved the informational value of protests to Party leaders while limiting their ability to transfer information between members of the general public. Striking this balance has allowed the Party to contain the threat that protesters pose to the regime (Lorentzen, 2013, 2017).

The few activists who do organize on behalf of an explicit ideological agenda are swiftly punished by the Party. Dissidents, such as the founders of the China Democratic Party and the leaders of the Charter 08 movement, are either arrested or sent into exile, where they often fall out amongst themselves (Potter, 2011). Censors prevent the policy demands of these activists from being discussed, while Party propagandists accuse them of carrying water for hostile foreign forces (Wright, 2002). The CCP’s ability to preclude the formation of a viable alternative to its rule ensures that the potential opposition remains leaderless, fragmented, and low in valence.

The Party’s efforts to manipulate the incentives of discontented citizens present them with an uncomfortable choice. Citizens who remain determined to take action often present themselves as wronged loyalists engaging in ‘rightful resistance’ (O’Brien, 1996; O’Brien and Li, 2006), confine themselves to narrower issues, such as land expropriation or pollution (Deng and Yang, 2013; Guo, 2001), or disguise their mobilization strategy by protesting alone (Fu, 2017). The adjustments made by these protesters generally blunt the ideological edge of their demands and emphasize their dissatisfaction with the valence of low-level officials.

As a result, we might anticipate that most of the citizens who are willing to take part in collective action are making a decision that is driven by the valence of the government and, in particular, by perceptions of government competence. They may consider protesting even if their ideological preferences are largely the same as the policy positions put forward by the Party.

4. An empirical stochastic model

In this study, I use Schofield’s stochastic valence model to assess the motivations that drive political participation in China. Formally, the model $M (λ, β)$ has individual utility, which is determined by the expression

u_{ij} (x_{i}, z_{j}) = λ_{j} - \sum_{k = 1}^{ω} β_{k} | | x_{ik} - z_{jk} | |^{2} + ϵ_{ij} .

Here, $λ_{j}$ is the exogenous valence of party j, and $β_{k}$ is a vector of positive ideological distances with length $ω$ , where $ω$ is the number of dimensions in the ideological space. $x_{ik}$ is individual i’s ideal point for the ideology dimension k, $z_{jk}$ is group j’s ideological position on dimension k, and $| | x_{ik} - z_{jk} | |$ is the Euclidean distance between the respondent $x_{ik}$ and the group $z_{jk}$ on dimension k. $ϵ_{ij}$ is an error term, which is assumed to follow the Type I extreme value or Gumbel distribution. This allows us to estimate the model in a multinomial logit framework.

$M (λ, β)$ is a pure spatial model which only incorporates terms for spatial distance and valence. It is also possible to specify a joint model $M (λ, θ, α, β)$ if we model individual decisions with additional terms for sociodemographic variables and attitudes toward the government. If we model individuals in this way, then utility for individual i is governed by the equation

u_{ij} (x_{i}, z_{j}) = λ_{j} + (θ_{j} \cdot ν_{i}) + (α_{j} \cdot τ_{i}) - \sum_{k = 1}^{ω} β_{k} | | x_{ik} - z_{jk} | |^{2} + ϵ_{ij} .

Here, $θ_{j}$ is a vector that contains the effect of each sociodemographic variable (age, education, sex, and family income) on the choice to join group j, while $ν_{i}$ is the vector of sociodemographic characteristics for individual i. The $(θ_{j} \cdot ν_{i})$ terms are scalar products, which we call the sociodemographic valences for group j.

$α_{j}$ is a vector that contains the effect of an attitudinal variable (the perception of government competence) on the choice of group j, while $τ_{i}$ is individual i’s score on the competence measure. We call the scalar product $(α_{j} \cdot τ_{i})$ the institutional valence for our model.

In traditional conceptions of the spatial model, political parties choose their ideological position to maximize their vote share. However, in China, the regime does not allow political groups to maximize their support by publicizing an ideological program. As a result, the ideology of a group is defined by the individual decisions made by its supporters. The ideology of each group for a given dimension k is defined to be the mean ideology of the group members on that dimension. If we suppose that each individual i chooses group $d_{i}$ , and $N_{j}$ is equal to the number of individuals who chose group j, then

z_{jk} = \frac{1}{N_{j}} \sum_{i | d_{i} = j} x_{ik} .

For both of our models, once we specify each group’s ideological position with $z$ , the probability that individual i chooses group j is

ρ_{ij} (z) = \Pr [[u_{ij} (x_{i}, z_{j}) > u_{il} (x_{i}, z_{l})], for all l \neq j] .

In other words, an individual chooses to join one group to maximize her own payoff. This payoff is based on the distance between her own ideology and the ideal point of each of the groups, which we have defined to be the mean of all of the group members’ ideal points. In the pure spatial model, this payoff is simply a function of the valence of each group and ideological distance, while in the joint model, this payoff is also shaped by an individual’s sociodemographic characteristics and her perceptions of government competence.

5. Empirical analysis

5.1. Data and measurement

The data for this study come from the Chinese Urban Governance Survey, which was administered in the summer of 2015 in 50 cities from 24 different provinces. This survey used GPS-assisted area sampling (Landry and Shen, 2005) to generate a nationally representative urban sample of 3513 respondents. Respondents answered a series of questions about political issues and the state’s capabilities in a variety of areas, ranging from its ability to maintain social stability to its ability to provide social welfare. Table 1 reports the essential descriptive statistics, as well as an index for perceived government competence, which is calculated as the mean of the seven measures of government capabilities.

Table 1.

Summary statistics.

Statistic	N	Mean	Standard deviation	Minimum	Maximum
Ideology (Left–Right)	3513	0.00	1.00	−3.81	4.03
Ideology (Authoritarian–Democratic)	3,513	0.00	1.00	−4.06	4.09
CCP member	3494	0.12	0.32	0	1
Age	3513	43.20	15.04	18	70
Education (Years)	3243	10.59	4.22	0	25
Female	3513	0.50	0.50	0	1
Family income (RMB)	1112	72,848	78,025	2500	800,000
Willingness to protest	2343	0.11	0.31	0	1
Propaganda capacity	3132	2.84	0.70	1	4
Stability capacity	3227	2.96	0.67	1	4
Price control capacity	3034	2.81	0.73	1	4
Tax collection capacity	2939	3.07	0.76	1	4
Representation capacity	3075	2.39	0.79	1	4
Social welfare capacity	3138	2.37	0.77	1	4
Control cadres capacity	3064	2.37	0.83	1	4
Competence	2662	2.70	0.53	1	4

Each respondent’s ideological position was estimated using Bayesian item response theory and a set of 12 ideology questions, which were selected to cover the most salient set of political, economic, and cultural issues in China. I used an ordinal model to take advantage of the full variation in the responses, which were given using a four-point scale (Quinn, 2004).

A two-dimensional model best fits the ideology data from the survey. The first dimension, which I call Left–Right, captures a divide between pro-market and anti-market sentiments. For the Left–Right dimension, the most discriminating questions ask respondents whether they believe that private ownership of property disadvantages working class people, whether privatization of state-owned enterprises should be outlawed, and whether state-owned enterprises should control key sectors of the economy.³ Respondents who agreed with these statements received negative scores on the first dimension, which would put them on the Left, while those who generally disagreed had positive scores, which are associated with the Right.

The second dimension of ideology is associated with political and cultural divisions. The items that load most heavily on this dimension include questions about whether Western multiparty democracy is suitable for China, whether freedom of speech will lead to chaos, and whether Confucianism is suitable for a modern Chinese society. Individuals who tended to agree with the statements received a negative score on the second dimension, which put them closer to the more authoritarian end of the spectrum; individuals who supported multiparty democracy and freedom of speech received positive scores on the second dimension.

Table 2 presents the difficulty and discrimination parameters for each ideology question. In general, the questions that load most heavily on a given dimension have discrimination parameters with large absolute values.

Table 2.

Difficulty and discrimination parameters.

	Difficulty Parameter	Left-Right Discrimination Parameter	Auth.-Dem.DiscriminationParameter
Private ownership of property disadvantages working class people 发展私有制经济会导致劳动⼈民沦为弱势群体	2.37	−1.09	0.20
Privatizing the assets of state-owned enterprises should not be allowed 不能允许民间资本兼并国有企业	1.75	−0.80	0.17
Attempting to control real estate prices will undermine economic development 试图控制房地产价格的⾏为会破坏经济发展	1.72	−0.71	0.13
Sectors important to people’s livelihoods must be controlled by state-owned enterprises 关系到国计民⽣的领域- 必须全部由国有企业掌控	1.51	−0.67	0.00
Marketization exacerbates economic inequality 市场化必然加剧贫富两极分化	2.07	−0.65	−0.01
Western Multiparty systems are unsuitable for China in its current state 西⽅的多党制不适合中国国情	2.78	−0.60	−1.04
Media should be allowed to represent the voice of specific social strata or interest groups 应当允许媒体代表特定阶层或利益集团发⾔	1.55	−0.53	−0.16
Indiscriminately imitating western-style freedom of speech will lead to social disorder 照搬西⽅式的⾔论⾃由- 社会就乱了	2.34	−0.50	−0.76
The modern Chinese society needs Confucianism 现代中国社会需要儒家思想	2.48	−0.46	−0.71
The minimum wage should be set by the state 最低⼯资应由国家规定	2.19	−0.44	−0.40
China’s current political system is the one that is best suited for China’s circumstances 我国⽬前的政治制度是最适合中国国情的	2.36	−0.34	−0.54
Individuals should be able to own land 个⼈应当可以拥有⼟地	1.88	−0.16	−0.14

To estimate the spatial model, respondents were divided into four mutually exclusive groups, based on whether they were members of the Communist Party and on whether they would be willing to consider protesting in the future. All respondents who were members of the Party are coded as CCP members, regardless of whether they said they would consider protesting. For the respondents who were not Party members, those who said they would never protest were coded as Bystanders, those who were unsure as Undecideds, and those who would consider protesting in the future as Potential protesters.⁴ A total of 19 respondents did not answer the survey question about membership in the Communist Party and were dropped from the analysis, leaving us with a total of 3494 observations.

Following previous research using the empirical stochastic model, the ideal point of each group is taken as the mean of the ideal points of the members of that group. Table 3 gives the proportion of our sample who fell into each group, as well as the mean ideal points for each group on both the Left–Right and the Authoritarian–Democratic dimensions of ideology. The left panel of Figure 1 presents the distribution of ideology estimates for the survey as a whole, while the right panel shows a close-up of the center of our distribution.

Table 3.

Group proportions.

Group	Respondents	Percentof sample	Mean Left–Right ideology	MeanAuthoritarian–Democraticideology
Bystanders	1821	52.1%	−0.046	−0.040
Undecideds	1049	30.0%	0.047	0.094
Potential protesters	215	6.2%	0.085	0.288
CCP members	409	11.7%	0.022	−0.205

Figure 1.

Ideological distribution of protesters and Party members in China.

6. Results

The results of our mixed logit model show that while both valence and ideological distance shape political participation decisions in China, valence is the more powerful factor. Institutional valence, or the perception of government competence, is the key factor that encourages individuals to consider protesting in the future, while sociodemographic valences, such as education, age, and income, are key factors that shape the decision to join the Communist Party.

Table 4 presents the results from the pure spatial model, $M (λ, β)$ . In this model, the spatial coefficients for both the Left–Right and Authoritarian–Democratic dimensions of ideology are significant. The valence terms are calculated with respect to the Bystanders, who are the highest valence group.

Table 4.

Pure spatial model results (base, Bystanders).

Variable	Coefficient (standard error)
Spatial distance
Left–Right ideology ( $β_{1}$ )	0.379** (0.181)
Authoritarian–Democratic ideology ( $β_{2}$ )	0.492*** (0.075)
Valence terms
Potential protesters $(λ_{Protesters})$	−2.137*** (0.073)
Undecideds $(λ_{Undecideds})$	−0.552*** (0.039)
CCP members $(λ_{CCP})$	−1.492*** (0.055)
Observations	3494
Log likelihood	−3899.469

p < 0.1; **p < 0.05; ***p < 0.01

I also estimate a joint model, $M (λ, θ, α, β)$ , which includes sociodemographic terms and a measure of institutional valence, the perceived competence of the government. To alleviate missing data concerns, I conducted the multinomial logit analysis after multiple imputation. Table 5 presents the results of the joint model after imputation, while Table A.1 in the Supplementary Material presents the results of the model if listwise deletion is used to deal with missing observations.

Table 5.

Joint model results after imputation (base, Bystanders)

Group	Variables	Coefficients (standard error)
	Left–Right ideology ( $β_{1}$ )	0.039 (0.190)
	Authoritarian–Democratic ideology ( $β_{2}$ )	0.293*** (0.079)
Potential protesters	Valence $(λ_{Protesters})$	−2.445 (1.723)
	Competence	−0.724*** (0.160)
	Age	−0.044*** (0.007)
	Education	0.015 (0.025)
	Female	−0.588*** (0.154)
	Log income	0.362** (0.157)
Undecideds	Valence $(λ_{Undecideds})$	1.860* (1.071)
	Competence	−0.038 (0.094)
	Age	−0.029*** (0.003)
	Education	−0.046*** (0.013)
	Female	−0.167** (0.080)
	Log income	−0.049 (0.096)
CCP members	Valence $(λ_{CCP})$	−9.742*** (1.166)
	Competence	−0.075 (0.119)
	Age	0.048*** (0.005)
	Education	0.214*** (0.021)
	Female	−0.591*** (0.119)
	Log income	0.368*** (0.092)
Observations		3494
Log likelihood		−3620.30

p < 0.1;**p < 0.05;***p < 0.01

The results show that spatial distances on the Authoritarian–Democratic dimension of ideology explain how individuals choose to protest or join the Party in China. Individuals do not, however, choose how they participate based on the Left–Right dimension of ideology, which concerns the proper role of the state in the economy. This result is robust to imputation and runs counter to our intuitions formed from similar analyses of democracies, where economic ideology is typically a strong factor that shapes vote choice (Schofield et al., 2011b,c).

The pre- and post-imputation models also come to the same conclusions when it comes to the issues of competence and valence. Individuals who do not perceive the government to be competent become significantly more likely to say that they would consider protesting in the future. Perceptions of government competence do not, however, determine whether individuals decide to join the Communist Party, or explain why some respondents are unsure when asked if they would protest in the future.

Once sociodemographic characteristics and perceptions of competence are taken into account, in both models the CCP members have a significantly lower exogenous valence⁵ as a group than the Bystanders, or the other groups. This suggests that something other than ideological distance, the perceived competence of the government, or sociodemographic characteristics explains why a relatively small share of the urban population opts to join the Communist Party. One possible explanation is that the Communist Party is not seeking to maximize its membership in the same way that parties in democracies seek to maximize votes, since it preferentially admits individuals from the more elite segments of Chinese society.

The model tries to account for this possibility by estimating the effect of sociodemographic valences on political participation. In the imputed sample, respondents are more likely to consider protesting in the future, instead of saying they would never protest, if they are younger, male, and, notably, more affluent, which disconfirms the hypothesis that the poor are the key potential opposition group.

Younger, male, and less well-educated respondents are more likely to say that they are unsure about protesting in the future, again in comparison with the people who would never protest. Conversely, respondents are more likely to be members of the Communist Party if they are older, male, better educated, and come from a higher-income family. The importance of sociodemographic valences for membership in the Communist Party is consistent with the argument that, in the reform era, the Party has strategically recruited elites and intellectuals, with less attention to their ideological bona fides.

Not all of these relationships are apparent in the pre-imputation sample. While the relationships between age and political participation are the same for all groups, the effects of sex, education, and family income on participation are only significant for Communist Party members.

7. Discussion

This paper has shown that the spatial model can help explain political participation, even when it is applied in a non-democratic context. Chinese citizens are more likely to consider protesting or joining the Communist Party if their ideological beliefs are consistent with the preferences of other people engaging in the same types of behavior. However, only the second, Authoritarian–Democratic, dimension of ideology produces this effect. Individuals in China do not choose to participate based on the Left–Right dimension, which captures their beliefs about the state’s role in the economy.

Valence explains more of the political participation decision than ideology. Individuals are willing to consider protesting if they take a dim view of the government’s competence, while sociodemographic characteristics, such as education, income, sex, or age, explain the decision to become a member of the Communist Party.

Of particular interest is the finding that, by the standards of a wide variety of democratic electoral systems, the ideological differences between the Communist Party and the members of the potential opposition are relatively small. This suggests that repressing ideological debates among elites and the media can forestall political polarization.

One of the biggest unanswered questions is whether this state of affairs is tenable if the Party loses control over political association. A well-known result in the political psychology literature is that members of deliberating groups tend to move toward a more extreme view than the pre-deliberation preferences of those individuals (Sunstein, 2002). This suggests that the narrow ideological distances between groups in Chinese society would not survive the onset of open political debate and freedom of association.⁶ If this comes to pass, political competition in China may turn on ideological differences after all.

Supplemental Material

JTP833190-APPENDIX – Supplemental material for A spatial valence model of political participation in China

Supplemental material, JTP833190-APPENDIX for A spatial valence model of political participation in China by Jason Y Wu in Journal of Theoretical Politics

Footnotes

Acknowledgements

I would like to thank James Adams, Claire Adida, Dan Chen, Bruce Dickson, Greg Distelhorst, Lei Guang, Steph Haggard, Seth Hill, Indriði Indriðason, Anna-Sophie Kurella, David Lake, Tianguang Meng, Molly Roberts, Sebastian Saiegh, Victor Shih, Susan Shirk, Michael Thies, Yiqing Xu, and Ning Zhang for their suggestions. Panel participants at APSA 2016, ACPS 2016, SCPI XII, and the Fudan-UC Young Scholars Conference also provided helpful feedback.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by the Smith Richardson Foundation, the Chiang Ching-Kuo Foundation, the 21st Century China Center, and the Carnegie Corporation of New York.

Notes

ORCID iD

Jason Y Wu

References

Adams

Merrill

(1999) Modeling party strategies and policy representation in multiparty elections: Why are strategies so extreme? American Journal of Political Science 43(3): 765.

Adams

Merrill

Grofman

(2005) A Unified Theory of Party Competition. New York, NY: Cambridge University Press.

Alvarez

Levin

Núñez

(2017) The four faces of political participation in Argentina: Using latent class analysis to study political behavior. The Journal of Politics 79(4): 1386–1402.

Ansolabehere

Snyder

(2000) Valence politics and equilibrium in spatial election models. Public Choice 103(3–4): 327–336.

CCP Central Committee (1981) On questions of Party history: Resolution on certain questions in the history of our Party since the founding of the People’s Republic of China. Beijing Review 24(27): 10–39.

Chen

(2017) Why do authoritarian regimes allow citizens to voice opinions publicly? The Journal of Politics 79(3): 792–803.

Deng

(1984) Selected Works of Deng Xiaoping, 1975–1982. Beijing: Foreign Languages Press.

Deng

Yang

(2013) Pollution and protest in China: Environmental mobilization in context. The China Quarterly 214: 321–336.

Dickson

(2014) Who wants to be a Communist? Career incentives and mobilized loyalty in China. The China Quarterly 217: 42–68.

10.

Dickson

Rublee

(2000) Membership has its privileges: The socioeconomic characteristics of Communist Party members in urban China. Comparative Political Studies 33(1): 87–112.

11.

Downs

(1957) An Economic Theory of Democracy. New York, NY: Harper & Row.

12.

(2017) Disguised collective action in China. Comparative Political Studies 50(4): 499–527.

13.

Groseclose

(2001) A model of candidate location when one candidate has a valence advantage. American Journal of Political Science 45(4): 862.

14.

Guo

(2001) Land expropriation and rural conflicts in China. The China Quarterly 166: 422–439.

15.

Hotelling

(1929) Stability in competition. The Economic Journal 39: 41–57.

16.

Kurella

Pappi

(2015) Combining ideological and policy distances with valence for a model of party competition in Germany 2009. Journal of Theoretical Politics 27(1): 86–107.

17.

Landry

Shen

(2005) Reaching migrants in survey research: The use of the global positioning system to reduce coverage bias in China. Political Analysis 13(1): 1–22.

18.

Lorentzen

(2013) Regularizing rioting: Permitting public protest in an authoritarian regime. Quarterly Journal of Political Science 8(2): 127–158.

19.

Lorentzen

(2017) Designing contentious politics in post-1989 China. Modern China 43(5): 459–493.

20.

Chu

Shen

(2016) Mass media, new technology, and ideology: An analysis of political trends in China. Global Media and China 1(1–2): 70-101.

21.

MacFarquhar

Schoenhals

(2006) Mao’s Last Revolution. Cambridge, MA: Harvard University Press.

22.

Mao

(1977) Selected Works of Mao Tse-Tung. vol. 5. Peking: Foreign Languages Press.

23.

McKelvey

Patty

(2006) A theory of voting in large elections. Games and Economic Behavior 57(1): 155–180.

24.

Micozzi

Saiegh

(2015) An empirical stochastic model of Argentina’s Impossible Game (1955–1966). Journal of Theoretical Politics 28(2): 266–287.

25.

Nathan

Shi

(1996) Left and Right with Chinese characteristics: Issues and alignments in Deng Xiaoping’s China. World Politics 48(4): 522–550.

26.

O’Brien

(1996) Rightful resistance. World Politics 49(1): 31–55.

27.

O’Brien

(2006) Rightful resistance in rural China. New York, NY: Cambridge University Press.

28.

Oliver

Ostwald

(2018) Explaining elections in Singapore: Dominant party resilience and valence politics. Journal of East Asian Studies 18(2): 129–156.

29.

Pan

(2018) China’s ideological spectrum. The Journal of Politics 80(1): 254–273.

30.

Potter

(2011) 4 June and Charter 08: Approaches to remonstrance. China Information 25(2): 121–138.

31.

Quinn

(2004) Bayesian factor analysis for mixed ordinal and continuous responses. Political Analysis 12(4): 338–353.

32.

Riker

Ordeshook

(1973) An Introduction to Positive Political Theory. Englewood Cliffs, NJ: Prentice-Hall.

33.

Schofield

(2007) The mean voter theorem: Necessary and sufficient conditions for convergent equilibrium. The Review of Economic Studies 74(3): 965–980.

34.

Schofield

Levinson

(2008) Modeling authoritarian regimes. Politics, Philosophy & Economics 7(3): 243–283.

35.

Schofield

Sened

(2005) Multiparty competition in Israel, 1988–96. British Journal of Political Science 35(04): 635–663.

36.

Schofield

Zakharov

(2009) A stochastic model of the 2007 Russian Duma election. Public Choice 142(1–2): 177–194.

37.

Schofield

Claassen

Gallego

et al . (2011a) Empirical and formal models of the United States presidential elections in 2000 and 2004. In: Schofield

Caballero

(eds.) Political Economy of Institutions, Democracy and Voting. New York, NY: Springer, pp. 217–258.

38.

Schofield

Gallego

Jeon

(2011b) Leaders, voters and activists in the elections in Great Britain 2005 and 2010. Electoral Studies 30(3): 484–496.

39.

Schofield

Gallego

Jeon

et al . (2012) Modelling elections in the Caucasus. Journal of Elections, Public Opinion & Parties 22(2): 187–214.

40.

Schofield

Jeon

Muskhelishvili

et al . (2011c) Modelling elections in post-communist regimes: Voter perceptions, political leaders and activists. In: Schofield

Caballero

(eds.) Political Economy of Institutions, Democracy and Voting. Berlin: Springer, pp. 259–301.

41.

Stokes

(1963) Spatial models of party competition. American Political Science Review 57(02): 368–377.

42.

Stokes

(1992) Valence politics. In: Kavanagh

(ed.) Electoral Politics. Oxford: Clarendon Press, pp. 141–164.

43.

Sunstein

(2002) The law of group polarization. The Journal of Political Philosophy 10(2): 175–195.

44.

Wright

(2002) The China Democracy Party and the politics of protest in the 1980s–1990s. The China Quarterly 172: 906–926.

45.

Meng

(2016) The nature of ideology in urban China. Presented at MPSA 2016.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.62 MB