Arseniy Samsonov
Welcome! My name is Arseniy Samsonov. I am an Assistant Professor of Economics at Ozyegin University in Istanbul.
I was previously a postdoc at the QSMS research group at the Budapest University of Technology and obtained my Ph.D. from UCLA in the Summer of 2021. My research areas are political economy and economic theory. Currently, I look at the role of information in politics from a game-theoretic point of view. My previous work used game theory and data analysis to study the welfare of ethnic minorities under different political regimes.
You may contact me at: arseniy.samsonov@ozyegin.edu.tr
My CV is here.
New! "Should Political Influence be Informed?" , joint with Maxim Senkov (European Research University).
Abstract:
Do voters benefit if politicians learn their needs? To answer this question, we provide a model where politicians compete by offering a public good to voters. Vot- ers differ in their ideological affinities to the politicians and the benefits the public good gives them. We compare a scenario where politicians target specific voters to one in which they only decide on the level of spending. The first scenario models informed politicians and the second reflects uninformed ones. When politicians are informed, and the budget is large, the incumbent politician wins by targeting a rel- atively small group of moderate supporters. When politicians cannot target specific voters, a large budget leads to generous promises that benefit voters. As a result, giving voter information to politicians leads to lower welfare. If the budget is low and politicians are informed, the Incumbent offers public goods to poor supporters of the Challenger. Winning them back would require the Challenger to drain the budget, so the Incumbent wins. Because poor citizens get the public good, overall welfare is higher than when politicians are uninformed.
Check out our recent working paper, "Information Agreements," joint with Kemal Kivanç Aköz (HSE).
Abstract: How transparent are informational institutions if their founders have to agree on the design? We analyze a model where several agents bargain over persuasion of a single receiver. We characterize the existence of an agreement that is beneficial for all agents relative to some fixed benchmark information structure, when the preferences of agents are state-independent, and provide sufficient conditions for general preferences. We further show that a beneficial agreement exists if, for every coalition of a fixed size, there is a belief that generates enough surplus for its members. Next, we concentrate on “agent-partitional” environments, where for each agent there is a state where the informed decision of the receiver benefits her the most. In these environments, we define endorsement rules that fully reveal all such “agent-states”. Endorsement rules are Pareto efficient when providing information at all agent-states generates enough surplus, and they correspond to a Nash Bargaining solution when the environment is also symmetric. Finally, we discuss the implications of our model in the context of political bargaining among the authoritarian elites over institutions and political parties over electoral campaigning in multiparty elections.
My paper "How can social media limit disinformation?" is submitted.
Abstract: Political disinformation is dangerous for democracies. Social and government pressure made Facebook and Twitter start labeling posts that contain disinformation. This policy became known as fact-checking. Under which conditions will social media platforms fact-check? I propose a model in which two platforms decide whether to fact-check a politician. One of the platforms is ex-ante more attractive to voters than the other. The politician chooses which platforms to use and how often to misrepresent the state of the world if it is bad for her. In equilibrium, fact-checking is more likely if the politician has low approval or one of the platforms is highly more attractive than its competitor. The model's policy implication is that reducing market power for the dominant social media firm can make fact-checking less likely.
My paper "The Fragmentation of views in a Democracy" is published in "Economics of Governance".
Abstract: Are voters in democracies more competent if there are more media outlets? To answer this question, I provide a game-theoretic model of media capture and political persuasion in democratic countries. In the model, there are two politicians, the Incumbent and the Challenger. They co-opt the media by offering them access to information. In exchange, the media support politicians who are available for interviews or include journalists in press pools. Voters choose like-minded media. I show that if the Incumbent is sufficiently popular and has little policy information, then media bias in her favor weakly increases in the number of media outlets. Otherwise, media bias in the Incumbent's favor weakly decreases in the number of media outlets. The welfare of voters weakly increases and decreases in the respective cases. The intuition is that, in equilibrium, the Incumbent can co-opt only one media outlet and ensure that enough voters read it. In this case, media outlets compete for access to the Incumbent and agree for a higher bias as their number increases.
My older working paper "Minorities in Dictatorship and Democracy" is available as a conference preprint and I am preparing it for submission.
Abstract: How does the level of democracy in a country affect the government’s treatment of ethnic minorities? I find that, on average, when the largest ethnic group in a society exceeds half of the population, ethnic minorities are treated better in autocracies and full democracies than in semi-democratic countries. The intuition is that under autocracy a leader needs little pop- ular support, and therefore a coalition of several minorities can rule. By contrast, in a semi-democracy, the leader needs the support of more peo- ple, so a coalition of small ethnic groups is insufficient; the largest group is enough and no other groups are necessary. Finally, highly democratic countries require broad support, and most ethnic groups get benefits. My model is based on the Baron-Ferejohn bargaining game and my empirical tests use the Ethnic Power Relations dataset.
You can find the do-files here.
"Persuasion in a Principle-Agent Model", with Noemie Cabau (QSMS), is available on SSRN and prepared for submission.
Abstract: We consider a principal-agent model in which the agent commits to a signal structure that generates messages based on his actions. There are also signals about the agent’s performance that he cannot control. The principal designs a payment scheme that depends on both types of signals. Both players are risk-neutral, and the negative payments only enter the agent’s payoff (e.g. criminal prosecution or daunting bureaucratic procedures). If the agent provides information, the principal can pay less to incentivize his effort. However, the agent may benefit because providing information can make punishment less likely. Because of this trade-off, the agent may choose an information structure such that he sometimes receives punishment when he exerts high effort. The model applies to a scenario where a bureaucrat must provide reports about his performance and is powerful enough to affect them.
"Logrolling and the power of single-mindedness", with Albert Sole Olle and Dimitrios Xefteris, is submitted and available on SSRN.
Abstract: We consider a committee that handles multiple issues, determining both the direction and extent of reforms for each, and investigate how logrolling affects outcomes. Our study provides a complete formal account of decentralized vote-trading in such a setting, by establishing two key findings: a) an essentially unique competitive equilibrium always exists, and b) logrolling improves the welfare of every committee member, and implements the Nash bargaining solution. With sufficiently many issues, a single-minded delegate fully dictates policy within her domain of interest, reflecting the disproportional influence of single-issue minorities in representative democracies.