Please use this identifier to cite or link to this item: https://repositorio.insper.edu.br/handle/11224/3861
Type: Tese
Title: Electoral competition and platform choice: a computational linguistics approach based on Brazilian data
Author: Pereira, Leila Albuquerque Rocha
Examination board: Novaes, Lucas Martins
Corbi, Raphael Bottura
Estevan, Fernanda Gonçalves de La Fuente
Bueno, Natália Salgado
Advisor: Firpo, Sergio Pinheiro
Academic degree: Doutorado
Publication Date: 2021
Abstract: This dissertation empirically investigates some valuable aspects of the electoral competition theories by taking advantage of computational linguistic methods that transform texts into data. I start by employing statistical text analysis methods (i.e., LDA model and Wordscores) to build and describe a unique dataset containing the topics discussed and the partisan-scores of Brazilian mayoral candidates' political platforms in 2012 and 2016 elections. Then, I use this dataset to test some theoretical implications of both the Downsian location models and the Salience Theory. From a Downsian perspective, I investigate the relationship between campaign spending and platform choice by testing whether a candidate's platform positioning strategy, compared to her opponent, is infuenced by the amount of money she can spend in an election. Then, from a Salience Theory perspective, I analyze the relationship between reputation and platform choice by estimating the incumbency effect on a candidate's platform issue concentration. In a nutshell, I find that candidates seem to be making strategic decisions concerning their platform's contents and partisan leaning. The data's descriptive analysis shows a systematic correlation between the candidate's characteristics and her platform's contents. Moreover, the empirical evidence supports the theories that connect platform choice to reputation but not the theories that establish a link between platform choice and campaign spending. In particular, I find that opposing candidates' strategies to diferentiate their platforms do not change significantly depending on the size of their campaign expenditures. Furthermore, I show that incumbents tend to produce more concentrated platforms than challengers.
Keywords (english terms): Electoral competition
political platforms
Downsian Models
Salience theory
natural language processing
Latent Dirichlet Allocation
wordscore
discontinuos regression
Language: Inglês
CNPq Area: Ciências Sociais Aplicadas
Copyright: O INSPER E ESTE REPOSITÓRIO NÃO DETÊM OS DIREITOS DE USO E REPRODUÇÃO DOS CONTEÚDOS AQUI REGISTRADOS. É RESPONSABILIDADE DOS USUÁRIOS INDIVIDUAIS VERIFICAR OS USOS PERMITIDOS NA FONTE ORIGINAL, RESPEITANDO-SE OS DIREITOS DE AUTOR OU EDITOR
Appears in Collections:Doutorado em Economia dos Negócios

Files in This Item:
File Description SizeFormat 
Leila_A_R_Pereira - Dissertation.pdfLeila_A_R_Pereira - Dissertation5.21 MBAdobe PDFThumbnail
View/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.