Are the polls rigged?

• Grant number: 17/12285-0
• Support Opportunities: Scholarships abroad - Research Internship - Master's degree
• Start date: September 04, 2017
• End date: December 11, 2017
• Field of knowledge: Applied Social Sciences - Economics - Economic Theory
• Principal Investigator: David Daniel Turchick Rubin
• Grantee: Felipe Ricardo Durazzo
• Supervisor: Filipe Robin Campante
• Host Institution: Faculdade de Economia, Administração e Contabilidade (FEA). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Institution abroad: Harvard University, Cambridge, United States
• Associated to the scholarship: 16/18987-4 - To manipulate or not manipulate, that is the question, BP.MS

Abstract

The effects of electoral polls in election outcomes is a studied topic in economics. However, to the best of our knowledge, the behaviour of pollsters have not yet been studied from an economic standpoint. This research project will study the incentives that pollsters have regarding the release of polls results. Using a two-candidate costly voting model, we assume that the pollster has a private information about which state has nature played, but it may report a false information to the public: if this happens, we say that the pollster manipulated the poll. An often heard criticism of pollsters is that they may manipulate a poll in order to benefit some candidate. This project may show that they have incentives to manipulate a poll even in the absence of ideological motives. Not necessarily a manipulated poll is bad: preliminary results show there are situations in which some degree of manipulation improves expected social welfare. (AU)