Sign in

Xu Xu

@xuxupolitics.bsky.social
118 followers 99 following 52 posts

Assistant Professor of Politics and International Affairs, Princeton University xu-xu.net

PostsRepliesMedia
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
11/We are grateful to the editors , reviewers, and many colleagues for their helpful comments, as well as our research assistants for their excellent work! Full paper here: doi.org/10.1177/0010...
doi.org
Sage Journals: Discover world-class research
Subscription and open access journals from Sage, the world's leading independent academic publisher.
000 0
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
10/ This is aggregate evidence, not a claim about every applicant. Many people pursue civil-service jobs out of public-service motivation, professional commitment, and a genuine desire to serve. Our argument is about how state power shapes incentives at scale.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
9/ CGSS data suggest that the inflow did not come mainly from the talent pool for other state agencies, but more likely from the private sector. This pattern echoes a broader dynamic often described as “the state advances, the private sector retreats” (国进民退) in China.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
8/After the VAT reform, tax-agency jobs in historically more corrupt prefectures attracted more—and higher-ability—applicants, suggesting that talent was not just chasing job security and promotion prospects, but the private returns to state power.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
7/The biggest increase was for tax jobs with real regulatory power—inspection, auditing, and enforcement. Applicants were especially drawn to positions where expanded state power was most directly exercised.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
6/After the reform, tax-agency jobs in high-exposure prefectures saw: +20.7% applicants; +18.4% applicant-to-recruit ratio; +1.1% interview cutoff scores. More people applied, and the applicant pool got stronger.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
5/ Using 166,012 civil-service job openings from 2010–2021, we compare prefectures more vs. less exposed to the VAT reform. Where the tax agency gained more power, its jobs became more popular, more competitive, and attracted stronger applicants.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
4/ We use China’s 2016 VAT reform, which made tax evasion harder by allowing input deductions and cross-checking invoices across firms. The reform strengthened the tax agency’s power over firms. We ask: when a state agency gains power, does talent move toward it?
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
3/In non-democracies, state building often means expanding control over society and the private sector. As power and resources become concentrated in the state, government jobs offer not only security but also access, authority, and opportunity—drawing talent into the state sector.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
2/ We study China’s “civil servant fever”: millions take the civil service exam each year. In 2021, one position drew 20,602 applicants for a single slot. The fever emerged after 2003, coinciding with state power expansion amid China’s growing private economy.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
Why do talented people compete so fiercely for government jobs in China? In our @cpsjournal.bsky.social paper with Bo Feng, Qiwei He, and Xin Jin, we identify state power as a key reason—with implications beyond China. doi.org/10.1177/0010... 🧵
120 2
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
11/We are grateful to the editors , reviewers, and many colleagues for their helpful comments, as well as our research assistants for their excellent work! Full paper here: doi.org/10.1177/0010...
doi.org
Sage Journals: Discover world-class research
Subscription and open access journals from Sage, the world's leading independent academic publisher.
000 0
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
10/ This is aggregate evidence, not a claim about every applicant. Many people pursue civil-service jobs out of public-service motivation, professional commitment, and a genuine desire to serve. Our argument is about how state power shapes incentives at scale.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
9/ CGSS data suggest that the inflow did not come mainly from the talent pool for other state agencies, but more likely from the private sector. This pattern echoes a broader dynamic often described as “the state advances, the private sector retreats” (国进民退) in China.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
8/After the VAT reform, tax-agency jobs in historically more corrupt prefectures attracted more—and higher-ability—applicants, suggesting that talent was not just chasing job security and promotion prospects, but the private returns to state power.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
7/The biggest increase was for tax jobs with real regulatory power—inspection, auditing, and enforcement. Applicants were especially drawn to positions where expanded state power was most directly exercised.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
6/After the reform, tax-agency jobs in high-exposure prefectures saw: +20.7% applicants; +18.4% applicant-to-recruit ratio; +1.1% interview cutoff scores. More people applied, and the applicant pool got stronger.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
5/ Using 166,012 civil-service job openings from 2010–2021, we compare prefectures more vs. less exposed to the VAT reform. Where the tax agency gained more power, its jobs became more popular, more competitive, and attracted stronger applicants.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
4/ We use China’s 2016 VAT reform, which made tax evasion harder by allowing input deductions and cross-checking invoices across firms. The reform strengthened the tax agency’s power over firms. We ask: when a state agency gains power, does talent move toward it?
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
3/In non-democracies, state building often means expanding control over society and the private sector. As power and resources become concentrated in the state, government jobs offer not only security but also access, authority, and opportunity—drawing talent into the state sector.
100 1
Xu Xu @xuxupolitics.bsky.social · 22/06/2026
2/ We study China’s “civil servant fever”: millions take the civil service exam each year. In 2021, one position drew 20,602 applicants for a single slot. The fever emerged after 2003, coinciding with state power expansion amid China’s growing private economy.
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
8/ We are grateful to the editors, reviewers, and many colleagues for their helpful comments, as well as our research assistants for their excellent work!
000 0
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
7/ As LLMs power search, assistants, and apps, censorship can become less visible—not just blocked links, but “polite” refusals or plausible-sounding falsehoods. That has implications for information access and global AI influence.
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
6/ This is observational (not causal), but the patterns are consistent with censorship via regulation and delegated enforcement. ChatGLM refuses far less than later China models released after regulation; non-China models change little across the same period.
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
5/ Language and training data matter, but they’re not the whole story: all models refuse more often in Chinese than in English. Still, the China vs. non-China gap is much larger than the Chinese vs. English gap within the same model.
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
4/ This gap shrinks on 30 less-sensitive prompts (China models’ refusal rates drop sharply), suggesting the differences aren’t fully explained by general capability gaps, product design choices, or market preferences.
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
3/ When they do answer, China-originating models give shorter and more inaccurate responses on those China-politics prompts—often failing to convey key components of the correct answer.
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
2/ We compare widely used, foundation LLMs developed in China and outside China (tested in 2023 and again in 2025) on 145 China-politics questions. China-originating models show much higher refusal rates—either no response or replies saying they can’t answer.
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
China’s chatbots are censored by the state. In our @pnasnexus.org paper with @jenpans.bsky.social, we find substantially higher levels of political censorship in large language models (LLMs) originating from China than those developed outside China. doi.org/10.1093/pnas... 🧵
doi.org
Political censorship in large language models originating from China
Abstract. A growing body of research on large language models (LLMs) has identified various biases, primarily in contexts where biases reflect societal pat
143 4
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
6/ This is observational (not causal), but the patterns are consistent with censorship via regulation and delegated enforcement. ChatGLM refuses far less than later China models released after regulation; non-China models change little across the same period.🧵
000 0
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
5/ Language and training data matter, but they’re not the whole story: all models refuse more often in Chinese than in English. Still, the China vs. non-China gap is much larger than the Chinese vs. English gap within the same model.🧵
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
4/ This gap shrinks on 30 less-sensitive prompts (China models’ refusal rates drop sharply), suggesting the differences aren’t fully explained by general capability gaps, product design choices, or market preferences.🧵
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
3/ When they do answer, China-originating models give shorter and more inaccurate responses on those China-politics prompts—often failing to convey key components of the correct answer.🧵
100 1
Xu Xu @xuxupolitics.bsky.social · 17/02/2026
2/ We compare widely used, foundation LLMs developed in China and outside China (tested in 2023 and again in 2025) on 145 China-politics questions. China-originating models show much higher refusal rates—either no response or replies saying they can’t answer.🧵
200 2
Xu Xu @xuxupolitics.bsky.social · 07/10/2025
Thank you!
000 0
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
9/ We are grateful to the editors, reviewers, and many colleagues for their helpful comments, as well as our research assistants for their excellent work! Read the full paper here: 🔗 doi.org/10.1086/734267
doi.org
Disguised Repression: Targeting Opponents with Nonpolitical Crimes to Undermine Dissent | The Journal of Politics
Why do authoritarian regimes charge political opponents with nonpolitical crimes when they can levy charges directly related to opponents’ political activism? We argue that doing so disguises politica...
010 1
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
8/ Our work shows that framing repression can transform it from something widely opposed to something the public may support. It also points to the central role of morality in understanding mobilization and repression.
110 1
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
7/ Non-political charges are associated with a decreased willingness of supporters to engage in dissent on behalf of the arrested individuals as well as decreased overall support for the critic.
110 1
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
6/ Analyzing over 3.6M Weibo posts from 2010 to 2014, we find that dissidents with larger online followings are more likely to be charged with non-political crimes during the 2013 crackdown on online critics in China.
120 2
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
5/ Disguised repression also deters other activists by blurring the boundaries of punishable actions and raising the risk that even minor or past wrongdoings can be used as pretexts for punishment, fostering fear and self-censorship.
110 1
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
4/ It appears that disguised repression demobilizes followers by undermining dissidents’ moral authority: Respondents perceive the dissident as less moral when they are charged with a nonpolitical crime as opposed to a political crime.
110 1
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
3/ To answer this question, we first conducted a survey experiment in China. The results shows that disguised repression—compared to blatant repression—reduces support for dissidents, lowers willingness to dissent on their behalf, and increases support for their repression.
110 1
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
2/ Authoritarian governments convict some dissidents using nonpolitical charges like corruption, tax evasion, or sex crimes, what we call disguised repression. But why do this when political activism is already illegal and others are punished with blatant repression?
110 1
Xu Xu @xuxupolitics.bsky.social · 06/10/2025
Why do authoritarian states charge political opponents with non-political crimes? In our @thejop.bsky.social paper with Jennifer Pan & @yiqingxu.bsky.social, we examine how *Disguised Repression* undermines opponents’ moral authority and mobilization capacity. doi.org/10.1086/7342...
25028 50
Xu Xu @xuxupolitics.bsky.social · 27/05/2025
8) A heartfelt thank you to the editorial team, the anonymous reviewers, and the many scholars—all of whom provided constructive feedback that greatly improved our paper! A special thanks to our brilliant research assistants for their great help!
010 1
Xu Xu @xuxupolitics.bsky.social · 27/05/2025
7/ This work demonstrates that digital technology has fundamentally changed not only propaganda content, but also how propaganda materials are produced and disseminated. Read the full paper here: 🔗 onlinelibrary.wiley.com/doi/10.1111/...
onlinelibrary.wiley.com
American Journal of Political Science | MPSA Journal | Wiley Online Library
The rise of social media in the digital era poses unprecedented challenges to authoritarian regimes that aim to influence public attitudes and behaviors. To address these challenges, we argue that au...
100 1
Xu Xu @xuxupolitics.bsky.social · 27/05/2025
6/ User engagement (video likes, comments, & reshares) is higher for central-level videos repurposed from local content compared to those originating directly from the center.
100 1
Xu Xu @xuxupolitics.bsky.social · 27/05/2025
5/ Using a deep-learning based, frame-to-frame video-similarity learning framework to compare millions of video pairs at different levels, we find that content often flows bottom-up: central accounts pick up and elevate videos produced by local accounts.
100 1
Xu Xu @xuxupolitics.bsky.social · 27/05/2025
4/ What do these accounts post? Very little is ideological or leader-focused. The majority of content portrays a moral society where ordinary people & officials do good deeds, aligned with “positive energy” (正能量). Gov content drastically differs from non-gov trending videos.
100 1
Xu Xu @xuxupolitics.bsky.social · 27/05/2025
3/ Massive #’s of regime-affiliated accounts are mobilized to produce and post videos on a daily basis, especially during weekdays.
100 1