Georgiy (Gosha) Syunyaev

The R package CausalQueries can be used to make, update, and query causal models defined on binary nodes. Users provide a causal statement of the form X -> M <- Y; M <-> Y which is interpreted as a structural causal model over a collection of binary nodes. Then CausalQueries allows users to (1) identify the set of principal strata—causal types—required to characterize all possible causal relations between nodes that are consistent with the causal statement (2) determine a set of parameters needed to characterize distributions over these causal types (3) update beliefs over distributions of causal types, using a stan model plus data, and (4) pose a wide range of causal queries of the model, using either the prior distribution, the posterior distribution, or a user-specified candidate vector of parameters.

@article{tietz2026causalqueries,
  title={Making, Updating, and Querying Causal Models using CausalQueries},
  author={Tietz, Till and Medina, Lily and Syunyaev, Georgiy and Humphreys, Macartan},
  journal={Journal of Statistical Software},
  volume={117},
  number={1},
  year={2026},
  doi={https://doi.org/10.18637/jss.v117.i01},
}

Classic arguments about federalist governance emphasize an informational or learning role for decentralizing policy authority, but in practice, ideological outcomes frequently motivate this choice. We examine the role of ideology in the allocation of policy-making power by modeling a two-period interaction between an elected central executive and two local governments. Decentralization reduces the executive’s ability to set policy and control externalities but potentially insures against future policy reversals. In this environment, partial decentralization is a common outcome. Complete decentralization arises when executives are unlikely to be re-elected, party polarization is high, and institutional hurdles to policy-making are big. These results help to clarify existing cross-national empirical findings on the determinants of centralization.

@article{syunyaev2022federalism,
  title={Decentralization and Ideology},
  author={Wilke, Anna M and Syunyaev, Georgiy and Ting, Michael M},
  journal={Political Science Research and Methods},
  volume={14},
  number={1},
  pages={231--239},
  year={2025},
  publisher={Cambridge University Press},
  doi={https://doi.org/10.1017/psrm.2025.13},
  url={https://www.cambridge.org/core/journals/political-science-research-and-methods/article/decentralization-and-ideology/1A81048F21CBF73FC24F862D89E1D913},
}

Widespread acceptance of COVID-19 vaccines is crucial for achieving sufficient immunization coverage to end the global pandemic, yet few studies have investigated COVID-19 vaccination attitudes in lower-income countries, where large-scale vaccination is just beginning. We analyze COVID-19 vaccine acceptance across 15 survey samples covering 10 low- and middle-income countries (LMICs) in Asia, Africa and South America, Russia (an upper-middle-income country) and the United States, including a total of 44,260 individuals. We find considerably higher willingness to take a COVID-19 vaccine in our LMIC samples (mean 80.3%; median 78%; range 30.1 percentage points) compared with the United States (mean 64.6%) and Russia (mean 30.4%). Vaccine acceptance in LMICs is primarily explained by an interest in personal protection against COVID-19, while concern about side effects is the most common reason for hesitancy. Health workers are the most trusted sources of guidance about COVID-19 vaccines. Evidence from this sample of LMICs suggests that prioritizing vaccine distribution to the Global South should yield high returns in advancing global immunization coverage. Vaccination campaigns should focus on translating the high levels of stated acceptance into actual uptake. Messages highlighting vaccine efficacy and safety, delivered by healthcare workers, could be effective for addressing any remaining hesitancy in the analyzed LMICs.

@article{solis2021covid,
  title={COVID-19 Vaccine Acceptance and Hesitancy in Low- and Middle-Income Countries},
  author={Sol{\'\i}s Arce, Julio S and Warren, Shana S and Meriggi, Niccol{\`o} F and Scacco, Alexandra and McMurry, Nina and Voors, Maarten and Syunyaev,  Georgiy and Malik, Amyn Abdul and Aboutajdine, Samya and Adeojo, Opeyemi and others, 64},
  journal={Nature Medicine},
  volume={27},
  number={8},
  pages={1385--1394},
  year={2021},
  publisher={Nature Publishing Group},
  doi={https://doi.org/10.1038/s41591-021-01454-y},
  url={https://www.nature.com/articles/s41591-021-01454-y},
}

The paper explores various measures of institutional quality in Russian regions, and compares those measures to each other. Such analysis leads to the conclusion that Russian regional institutions are essentially multidimensional, and therefore comparisons of Russian regions in terms of their overall institutional quality could be problematic. New institutional indices are derived from Russian enterprise surveys held under the BEEPS project of the European Bank of Reconstruction and Development. Such indices yield a typology of Russian regions in terms of efficacy of regional administrations’ control over economy and bureaucracy in their regions. Dynamics of regional institutional indices is investigated against the backdrop of Russia-wide institutional trends.

@article{baranov2015not,
  title={How (not) to Measure Russian Regional Institutions},
  author={Baranov, Alexey and Malkov, Egor and Polishchuk, Leonid and Rochlitz, Michael and Syunyaev, Georgiy},
  journal={Russian Journal of Economics},
  volume={1},
  number={2},
  pages={154--181},
  year={2015},
  publisher={Elsevier},
  doi={https://doi.org/10.1016/j.ruje.2015.11.005},
  url={https://www.sciencedirect.com/science/article/pii/S2405473915000227},
}

We provide a theory and empirical evidence indicating that the rotation of ruling elites in conjunction with elites’ asset ownership could improve property rights protection in non-democracies. The mechanism that upholds property rights is based on elites’ concern about the security of their own asset ownership in the event they lose power. Such incentives provide a solution to the credible commitment problem in maintaining secure property rights when institutional restrictions on expropriation are weak or absent.

@article{polishchuk2015ruling,
  title={Ruling Elites' Rotation and Asset Ownership: Implications for Property Rights},
  author={Polishchuk, Leonid and Syunyaev, Georgiy},
  journal={Public Choice},
  volume={162},
  number={1},
  pages={159--182},
  year={2015},
  publisher={Springer},
  doi={https://doi.org/10.1007/s11127-014-0210-2},
  url={https://link.springer.com/article/10.1007/s11127-014-0210-2},
}

@inproceedings{polishchuk2015property,
  title={Property Rights without Democracy: The Role of Elites' Rotation and Asset Ownership},
  author={Polishchuk, Leonid and Syunyaev, Georgiy},
  booktitle={Elites, Institutions and the Quality of Government},
  editor="Dahlstr{\"o}m, Carl and W{\"a}ngnerud, Lena",
  pages={167--186},
  year={2015},
  publisher={Palgrave Macmillan, London},
  doi={https://doi.org/10.1057/9781137556288_10},
  url={https://link.springer.com/chapter/10.1057/9781137556288_10},
}

We study the impact of Russian regional governors’ rotation and their affiliation with private sector firms for the quality of investment climate in Russian regions. A theoretical model presented in the paper predicts that these factors taken together improve “endogenous” property rights under authoritarian regimes. This conclusion is confirmed empirically by using Russian regional data for 2002—2010; early in that period gubernatorial elections had been canceled and replaced by federal government’s appointments. This is an indication that under certain conditions government rotation is beneficial for economic development even when democracy is suppressed.

@article{syunyaev2014climate,
  title={Investment Climate and Government Turnover in Russian Regions},
  author={Syunyaev, Georgiy and Polishchuk, Leonid},
  journal={Voprosy Ekonomiki (In Russian)},
  number={2},
  pages={88--117},
  year={2014},
  doi={https://doi.org/10.32609/0042-8736-2014-2-88-117},
  url={https://www.vopreco.ru/jour/article/view/616?locale=en_US},
}

Many autocrats use state-owned media to claim credit or shift blame for policy outcomes, often by shaping perceptions of whether central or local government is responsible. Are such strategies effective when citizens know the media is government-controlled? I present evidence from a survey experiment with over 4,000 Russians randomly assigned to watch news reports from the state-owned channel Rossia-1. The reports highlighted the central government’s monitoring of road maintenance and natural disaster management—both local responsibilities. Exposure did not shift beliefs about who is responsible, but it improved perceptions of policy performance and increased support for government. These results fit a Bayesian learning model in which citizens, aware of media bias and true responsibility, infer that the center associates itself with a policy only when performance is good. Propaganda can thus persuade precisely because citizens know outlets are state-controlled—an effect strongest among those who consume less state media and are more skeptical of it—providing further evidence for the Bayesian learning interpretation.

Open and free media are often seen as crucial for political accountability, particularly in weak democracies and autocracies where they can catalyze regime change. Existing evidence supports this view but typically assumes that independent outlets are well known and trusted by the public. This paper tests that assumption. I field a panel 2x2 factorial experiment in Novosibirsk, Russia’s third-largest city, exposing respondents to pre-recorded local news reports on a highly salient issue—public healthcare provision. Despite high treatment compliance, exposure to independent local news had no meaningful effect on beliefs about policy performance, responsibility attribution, or government evaluations. I also find no evidence of treatment effect heterogeneity across key pre-treatment characteristics. These findings suggest that the accountability-enhancing effects of independent media documented in prior work may be limited to outlets with an established reputation and audience.

State propaganda in autocracies often creates information bubbles, leaving citizens unwilling or unable to switch to alternative sources. This study shows that propaganda’s sway can be decreased even without escaping such bubbles. In a panel experiment with 1,176 Russian citizens, we randomly assigned participants to analyze the reporting patterns of (i) propagandistic state-owned TV content only or (ii) both state-owned and balanced non-governmental TV content over four weeks. Two weeks after the intervention, both treatments reduced support for the authoritarian regime but did not undermine trust in or exposure to state media. Exploiting the differences in news channel composition across treatments, we find support for two mechanisms explaining these effects: increased attention to familiar media content and increased consumption of news from non-government sources. Subgroup analyses show that these mechanisms operate primarily among citizens who already exhibit skepticism toward the regime and propaganda.

Internet access is expanding in low- and middle-income contexts, yet its political effects remain unclear and may depend on how people connect. We study a pre-registered field experiment in a low-income South African township, where the high cost of mobile data constrains use. Around the 2024 national election, we randomly provided households with 8–12 months of uncapped home internet and cross-randomized exposure to an online get-out-the-vote campaign. Combining household surveys, verified turnout, and behavioral traces, we find that home connections increase in-person local political engagement—attendance at ward/community meetings and complaints to ward/municipal officials—while leaving voter turnout and online political participation unchanged. We detect no systematic shifts in political knowledge, views, or broad news consumption patterns. Evidence is consistent with a behavioral channel in which home connectivity raises time spent at home and incidental exposure to local governance, with suggestively stronger effects among men. The results show that expanding home internet, distinct from mobile data, can strengthen place-based participation without mobilizing online activism or turnout, underscoring that the mode of access conditions the political consequences of going online.

Refugee integration is a central policy challenge. Expanding on prior research on policy debates targeting immigrants, we examine whether host-country debates about military aid to refugees’ homeland shape their psychological integration. In a survey experiment with 2,631 Ukrainian refugees in Germany, participants viewed authentic statements by German politicians expressing varying levels of support for military aid. On average, these messages had no detectable effect. Yet this null masks important heterogeneity. By linking our experiment to daily variation in national news coverage, we find that when conflict salience is low, exposure to any political statement on military aid—supportive or opposing—reduces psychological integration. These effects vanish when the issue is already prominent in the media. We argue that this pattern reflects a contestation mechanism: refugees interpret the very visibility of political debate, rather than its specific content, as a signal that host-country support is contested and potentially less secure.

In response to Russia’s 2022 full-scale invasion of Ukraine, multiple citizen-led communication campaigns leveraged modern information technology to transmit millions of messages aimed at providing uncensored information to Russians via emails, texts, social media apps, and website commentary. But can such initiatives penetrate authoritarian censorship and propaganda, and does success depend on the types of messaging employed? To investigate, we partnered with Mail2Ru, a prominent citizen-led grassroots initiative, to conduct two pre-registered field experiments on the effectiveness of citizen-to-citizen email communication. We implemented a “mega-study,” inviting behavioral science experts to design theoretically informed persuasive messages encouraging Russians to watch a video about the war produced by an independent news organization. Approximately 260,000 email recipients were randomly assigned one of eleven treatments or two control messages. We report four key findings. First, information technology’s scalability allows even poorly designed messages to reach and engage large audiences with uncensored information. Second, highlighting both the limitations and potential of citizen-led initiatives, nearly all expert-designed messages failed to outperform controls, but our top-performing message nearly doubled engagement with the video, a result successfully replicated in a second experiment. Third, effectiveness was strongly correlated with “conversational receptiveness,” the conveying of thoughtful engagement with opposing perspectives. Fourth, exploratory analyses offer suggestive evidence that messages from female senders are more effective. These insights into which messaging strategies can meaningfully increase engagement with uncensored information offer evidence-based guidance for citizen-led efforts to counter authoritarian propaganda.

This project uses meta-analysis tools to study the relative performance of sampling and estimation strategies to measure the prevalence of hard-to-reach populations. Using data from 8 field studies of human trafficking in developing countries, we estimate the relative performance of the most common measurement strategies. These include successive sampling and link-tracing methods based on respondent-driven sampling, network scale-up methods based on time-location, and proportional sampling. Scholars can use the framework we provide for diagnosing their study designs and rolling analysis of the relative performance of measurement strategies.

A multi-country panel experiment on how people judge which news sources to trust and whether those judgments can be improved. Building on our panel experiment in Russia, which showed that directing attention to what and how state media report reduces the persuasiveness of propaganda, we field one harmonized design in the United States, Russia, Kazakhstan, Turkey, Tanzania, Brazil, and India. We measure how accurately people recognize which outlet reports what, how they revise their trust in sources after examining their coverage, and whether an intervention that directs attention to how sources report improves these judgments across media environments, from pluralistic democracies to tightly controlled autocracies.

Personalist autocracies fuse the ruler’s public persona with the political body of the regime. We argue that visible signs of leader weakness—whether physical frailty or public defiance by other political actors—serve as cues of diminished regime stability by altering citizens’ expectations about the regime’s capacity and durability. Declining perceptions of stability should then have politically asymmetric consequences. Among regime supporters, they should reduce feelings of security and heighten anxiety, weakening political support. Among regime opponents, they should increase political efficacy and perceived control, making collective action appear more feasible. We will test these claims in an online survey experiment with Russian respondents. Participants will be randomly assigned to one of four brief news-style reports about public events involving the leader or to a no-report control. In the treatment conditions, the accompanying video depicts either physical weakness or a public challenge from another political actor. The paired comparison conditions hold the report text and event context constant while presenting a same-event video excerpt without the focal weakness content: no salient physical-weakness episode in the appearance pair and no public challenge in the criticism pair. The appearance contrast therefore estimates the effect of encountering a naturalistic, bundled episode of diminished vigor, not the effect of age alone. This design allows us to estimate whether distinct manifestations of leader weakness produce a common shift in perceived regime stability and to separate those effects from exposure to the underlying event. Our primary outcome is perceived regime stability. Downstream outcomes measure security and anxiety, political efficacy and perceived control, regime support, and self-reported willingness to engage in low- and moderate-cost collective action. We also examine whether responses vary by prior regime orientation and attitudinal strength. The study identifies a micro-level mechanism through which an autocrat’s public image may contribute to the resilience or collapse of personalist rule.

Georgiy (Gosha) Syunyaev

Publications

Working Papers

Work in Progress

Software