China Working Papers

Ongoing Studies:

China Policy Barometer: In this annual survey we ask Chinese citizens what they think about the major policies debates facing the nation.

China Policy Barometer

For more information, visit

Working Papers:

Blurring the Lines: Rethinking Censorship under Autocracy (with Li Shao and Charles Crabtree)

Abtract: Self-censorship is subtle and difficult to observe, but its effects are potent and pernicious. We argue that policies encouraging self-censorship not only make it easier to stifle critical viewpoints, they also disadvantage those on the periphery of politics and power. Our guiding intuition is that actors without a connection to the regime have an informational disadvantage about what is on and off limits, which causes them to self-censor to a greater extent than those with closer links to the regime. We explore this hypothesis through the case of China, one of the world’s most restrictive information environments and a sanctum of self-censorship. Leveraging evidence from virtual field experiments and online surveys, we show that self-censorship practiced by censors, journalists, and even average netizens is closely related to their relative proximity to the regime. Specifically, members of the private media and politically unconnected individuals self-censor to a greater extent than their public sector and state-affiliated peers. This surprising finding helps explain how China’s rulers have been able to successfully contain and control a mushrooming information sector with only moderate use of overt censorship. It also helps explain how and why China’s state-backed media, despite a reputation for pro-regime bias, remain surprisingly competitive in the domestic market.

In Public or in Private: Repressing Dissidents in China

Abstract: When do states repress in public and when do they do so private? I argue that states repress publicly when they believe that doing so will deter future dissidents from engaging in similar challenges against the state. By contrast, if the state believes that public repression is unlikely to deter future challengers, they prefer to repress in private. Leveraging a unique dataset on dissidents in China, I demonstrate the theory in action by showing that repression of political and welfare dissidents is more likely to occur publicly than is repression of religious and ethnic dissidents. The relationship is robust to regional and temporal controls as well as legal institutions. One implication of this study is that the public reputations associated with repressive regimes probably have more to do with the way regimes want to be perceived than with the actual levels of repression they engage in.

We Asked You: Policy Preferences and Consultative Democracy in China (with Sinan Chu)

Abstract: China’s leaders describe their system of rule as “consultative democracy”, whereby the public is forbidden from organizing on matters of politics but encouraged to participate on issues of policy. In particular, citizens are routinely solicited for input on upcoming policy debates. But who participates and how are their opinions incorporated into the policymaking process? We employ an online survey instrument designed to measure public preferences on a range of policies recently debated by China’s National People’s Congress (NPC). Pairing participation and policy preference measures with latent measures of ideology from our survey, and outside surveys, we extrapolate a predicted ‘public opinion’ for the broader Chinese constituency. Next, we compare final policy decisions, based on whether or not they were opened for public consultation, to estimate the predicted effect of public opinion on policy choices. Our findings show that consultation does influence policy choice and that the effects are larger on policy items for which the state reported larger citizen response rates. Using our main model, we also make out-of-sample predictions about legislative actions scheduled for the upcoming 2016 NPC meeting.

Crisis of Confidence: Measuring trust with experiments in China (with Li Shao)

Abstract: Existing studies suggest that Chinese citizens hold their governments in high esteem, notably higher than in prominent Western democracies. One prominent explanation is that Chinese citizens view the central government as an efficient and effective allocator of economic resources, evinced by China’s consistently high growth rates. There are two problems with this interpretation. First, measurement bias—social desirability and scaling—may be leading to inflated and incongruous measures of trust. Second, it is unclear what the link between faith in the government and faith in its capacity as an economic actor actually is. In this study, we address both challenges by (a) estimating levels of trust using a series of indirect and unbiased survey experiments, and (b) linking revealed trust to a costly economic choice that hinges on the belief that government is not only trustworthy but also an effective. Specifically, we will estimate a Chinese citizen’s willingness to make charitable donations conditional on their knowledge of whether or not the receiving organization is government run, privately run, or foreign-run, knowledge that will be randomly manipulated in the survey experiment. We find that respondents are willing to donate significantly larger funds to foreign and private organizations. Using additional survey information, we show that these effects are closely associated with concerns over government integrity.

Cosponsorship and Information Provision in China’s Local People’s Congress System (with, Steven Oliver).

Abstract: How can leaders in authoritarian regimes mitigate the Dictator’s Dilemma? Recent literature suggests legislatures offer one possible solution. Yet few studies identify specific micro-institutions through which deputies reveal information to leaders. We argue that rules that permit cosponsorship of legislative proposals by heterogeneous coalitions of deputies constitute one such institution. Moreover, we argue that information revealed by heterogeneous coalitions allows leaders to reduce uncertainty about the relationship between policy and outcomes and improve their chances of retaining office. We describe this argument in the context of China’s system of local People’s Congresses and provide supporting evidence from recent plenary sessions of the Shenzhen Municipal People’s Congress.

Network Supervision: Public Participation and Anti-Corruption in China

Abstract: Authoritarian regimes, especially those that lack regular elections, have notoriously weak bottom-up accountability and, as a result, are at higher risk of petty corruption. Modern technology offers a way to bridge these faults by giving citizens a low- cost alternative through which to inform disciplinary authorities when they observe official impropriety. In China, for instance, the Central Disciplinary and Inspection Commission has encouraged hundreds of thousands of citizens to report on government misconduct using online submission portals. How effective have these policies been? Have they resulted in more investigations? Are they affecting official behavior? I answer these questions by leveraging legal and technical proxies of citizen reporting capacity. Specifically, I find that citizen reporting, when buoyed by whistleblower protections, Internet penetration, and government website transparency, significantly predicts anti-corruption investigations, even after controlling for socio-economic variation.