Who Believes Government Propaganda? Experimental Evidence from China

• In Event: Experiments in Political Economy

Abstract

Who believes propaganda in non-democracies? There is a burgeoning stream of literature about the role of political propaganda in the fields of political economy and authoritarian politics (Yanagizawa-Drott 2014; Adena et al. 2015; Huang 2015; Cantoni et al. 2017; Huang Forthcoming; Peisakhin and Rozenas Forthcoming). However, most existing studies focus on whether propaganda works in general rather than who believes the propaganda. It remains an open question that which groups of citizens are more prone to influences of propaganda, especially when propaganda often misrepresents or distort the reality.

The critical challenge in discerning effects of propaganda based on survey data is “false believers” caused by the social desirability bias and expressive values of answering survey questions. Citizens under authoritarian rules may conceal or falsify their true opinions about propaganda in survey responses for different reasons. They may use the chance of responding survey questions to express their nationalist or pro-regime sentiments even though they are not persuaded and convinced by propaganda. Other false believers may appear to believe propaganda because they fear political. Most studies of contemporary political propaganda, however, take respondents’ answers to direct questions at face value and mistakenly count these “false positives” into those who genuinely believe the propaganda.

We offer the first systematic and individual-level demographic analysis of the effects of propaganda in a non-democracy based on a nationally representative urban resident survey in China. To address the problem, this paper uses the approach of incentivized responses; a technique that has recently been applied in the field of American politics (Bullock et al. 2015; Prior, Sood and Khanna 2015). By randomly distributing monetary incentives for correct responses and admitting ignorance when subjects answer factual questions, this approach helps us minimize insincere responses among the treated and identify genuine believers of propaganda. To further identify disguised believers who fear political coercion, we apply a second experimental manipulation. An orthogonal treatment group is asked whether they have heard of a quote by Xi Jinping, which explicitly warns against stigmatization of the revolutionary history of the Chinese Communist Party. To test the hypothesis that respondents may not bother to think thoroughly and thus opt for a prevailing default option that most people support, we also include an additional treatment in one of the propaganda questions: one of three response items is randomly assigned with the label that 80% respondents choose the very item.

The core survey questions are around eight factual questions that correspond to official propaganda of the Chinese government. They cover three major categories of political propaganda: praising the regime, taunting of foreign countries, and shirking responsibilities. In closed-ended questions, subjects will choose from an option that matches official propaganda, an option of correct response, a confounder, and a “don’t know” option. In open-ended questions (such as comparing GDP per capita of China and that of Taiwan), official propaganda explicitly suggests a biased answer (China is depicted richer than it actually is). We also add two factual questions that are not about propaganda, so we can confirm that monetary incentives and signals of coercion do not induce biased responses in other non-propaganda questions. Based on responses to a series of demographic and attitudinal questions, we examine demographic and behavioral variables that explain citizens’ beliefs in propaganda.

Authors

• Fengming Lu, Princeton University

• Milan Svolik, Yale University