Full text
| |
| Author(s): |
Mauricio Yoshida Izumi
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Faculdade de Filosofia, Letras e Ciências Humanas (FFLCH/SBD) |
| Defense date: | 2013-10-24 |
| Examining board members: |
Paolo Ricci;
Fernando de Magalhaes Papaterra Limongi;
Lúcio Remuzat Rennó Júnior
|
| Advisor: | Paolo Ricci |
| Abstract | |
Ideal point estimation methods based upon the spatial theory of voting has been broadly used to represent legislatures around the world. However, such methods were initially thought to analyze the U.S case, that has a bipartisan system and in which party switching are rare events. In countries like Brazil, some precautions must be taken. Thus, the rst aim of this study is methodological. As the assumptions of the error distribution are not feasible to the Brazilian case, the use of techniques like W-Nominate and IDEAL are not recommended. So, we argue in favor of the use of the non-parametric method Optimal Classication. Further, in this direction, we discuss the problem of the unit of analysis and how it can mislead the choice of the number of relevant dimensions. The second aim is substantive. What we intend to answer is how many and which dimensions are necessary to represent fairly well the preferences of Brazilian senators. We will show that only one dimension that manifest the conict between government and opposition is sucient. (AU) | |
| FAPESP's process: | 11/03153-7 - The determinants of legislative behaviour in the brazilian Senate (1989-2010) |
| Grantee: | Maurício Yoshida Izumi |
| Support Opportunities: | Scholarships in Brazil - Master |