Authoritarian regimes have increasingly leveraged foreign media to project influence within democracies, yet evidence of these co-opted outlets’ actual effects remains scarce. This study presents findings from a field experiment conducted during Taiwan’s 2020 general election, assessing the impact of The China Times, a Beijing-backed media conglomerate, on voter behavior and attitudes. The experiment incentivized participants to engage in sustained consumption of real-time news from this outlet in the weeks leading up to the election. Results from a panel survey linked to individual-level web-tracking data reveal that exposure to The China Times sways voters in favor of the People’s Republic of China (PRC). These effects, however, are primarily driven by nonpartisan and PRC-friendly voters. To the extent that I find effects among PRC-skeptics, they show evidence of backfiring. As Beijing’s media co-optation extends beyond Taiwan, my findings have broader implications for understanding the effectiveness and limitations of authoritarian influence operations.

This study investigates how perceptions of government responsiveness influence citizen compliance in authoritarian regimes, specifically examining how exposure to information about reported government responsiveness affects individuals’ willingness to pay taxes in China. Results from a survey experiment among prospective Chinese business elites show that respondents exposed to information about improved government responsiveness demonstrated significantly lower tax compliance than those in the control group. Qualitative evidence revealed that respondents often perceived the positive message as state propaganda inconsistent with their personal experiences, potentially triggering backlash. The negative effect was particularly pronounced among low self-monitors. The findings contribute to the literature on the backfire effects of hard propaganda, highlighting how the disconnect between citizens’ lived experiences and pro-regime messaging can undermine policy objectives rather than simply fail to persuade.

Are host populations more accepting of immigrants who are racially similar and/or linguistically proficient in the host vernacular? The empirical focus in the literature has been largely dominated by Western democracies where the host society is white—and therefore the immigrants are often non-white. As such, we lack a theoretical explanation for how race moderates other markers—for example, language— when it comes to immigrant attitudes. To remedy this, we shift the focus to Taiwan, where the “New Residents”—a new catch-all census category for all post-1987 immigrants regardless of race, language, and national origins— offers an empirical opportunity to test our theory. In a conjoint experiment of Taiwanese attitudes and a survey of New Residents, we find attitudes are (1) most positive for Han Chinese who can speak a Taiwanese vernacular; (2) the least positive for Han Chinese who cannot speak a Taiwanese vernacular; and (3) relatively positive when immigrants are neither racially similar n or linguistically proficient. These findings, however, are conditional on the New Residents being from a non-politicized country (i.e., not China). The results have implications for how we study immigration, Taiwanese politics, and the Chinese diaspora.

Governments can grant political concessions to induce quasi-voluntary compliance with taxation, yet empirical evidence probing the taxation–representation connection remains inconclusive. We contend that this association remains valid but it is primarily confined to business elites in nondemocratic regimes because the same wealth that exposes them to state predation also incentivizes them to endorse tax policies that offer greater political representation. We test our argument by evaluating preferences for hypothetical tax reforms in separate samples of business elites and ordinary citizens in China. We find that business elites show stronger preference than nonelites for tax policies that include advances in political representation. We explore various mechanisms for our results and find support for government credibility, tax ownership, and tax salience considerations.

While authoritarian regimes are often characterized by their civil liberty restrictions, some dictatorships acknowledge the ethnolinguistic diversity of their population. Are minorities in multiethnic authoritarian states more likely to trust the government when their language is recognized? In this paper, we argue while recognition of a group’s language improves trust in democracies through a substantive representation mechanism, the same cannot be said in authoritarian regimes. Instead, recognition is a mere symbolic gesture. Such window-dressing efforts call attention to the horizontal inequality between hegemon and minority groups—and such, minority language recognition is associated with negative political trust. We test our argument with the World Values Survey. By identifying which minority groups have been afforded linguistic recognition, we find evidence of a significant—but negative—link between recognition and political trust.

When and how does state indoctrination work? Building upon research on motivated reasoning and family socialization, I argue that only those individuals whose parents have connections to political patronage are subject to state indoctrination because their pro-regime biases transmitted from parents induce higher receptivity prior to government messages. Focusing on political education in China, I conduct a quasi-experimental analysis exploiting the sharp variation in textbook content generated by China’s most recent curriculum reform. Results based on a national survey show that the new politics textbooks successfully affected only those individuals whose parents had worked for the government. The finding survives extensive robustness checks and falsification tests. I also consider several alternative explanations of the effects: preference falsification, selective attention, parental indoctrination, and educational quality. This paper not only highlights the role of intergenerational transmission in moderating the effectiveness of state indoctrination but also casts doubt on the actual degree to which regimes can change minds by changing educational content.