Project description

In order to discover what was unique about Brazil’s president Jair Bolsonaro inauguration speech, we used quantitative text analysis techniques to discover which words were characteristic of him – that is, words that he used a lot and all the former presidents used less. We discovered that he was obsessed with terms linked to the word “ideology”: he uses the term to talk about the left-leaning opposition and vows to rule the country without an “ideological” agenda.

In the past, the key words were usually linked to more concrete issues. In 2003, presidente Lula spoke about “hunger” more than any other. Dilma Rousseff was concerned about oil and corruption. Fernando Henrique Cardoso talked about inequality and citizenship rights.

What makes this project innovative?

This data-driven story displays an objective measure of the central themes of Bolsonaro’s inauguration speech, which was a landmark in contemporary Brazilian history. The method we developed allowed us to compare him to former rulers as well. When doing this analysis for past governments, we were also able to compare the way each president presented itself actual policies implemented. It’s an interesting way of contrasting political discourse and practice when looking at our recent history.

What was the impact of your project? How did you measure it?

The piece was able to spark interesting social media conversations, specially regarding the apparent lack of interest that Bolsonaro showed for social justice issues in his speech. As well as our others text analysis projects, it received attention from researchers both in the language and political science areas.

Source and methodology

The analysis was made with the inaugural speeches of all Presidents since 1889. Until 2003, when Lula assumed the Planalto, the pronouncements were taken from the book "Palavra de Presidente", by researcher João Bosco Bezerra Bonfim, who has a master in linguistics at UnB (National University of Brasília). After that, the presidential inaugural speeches made at the National Congress were used. The transcription came either from the Presidential Library website or from the newspaper archives. Since Michel Temer never made an official inauguration speech to Congress, we used his first public pronouncement after Dilma Rousseff was removed from office due to her impeachment process. To determine which words are characteristic of each candidate, we made the following calculation: the number of times a word is used in a given speech divided by the number of times the word occurs in all other speeches. The resulting ratio determines how unique a term is. Words used exclusively in a single speech were ignored. A practical example: for every thousand words spoken by Lula, in his first term, the word "hunger" appeared approximately six times. In all other speeches, this ratio was lesser than one. This means, therefore, that the term is typical of him. The algorithm used to do the computation is available on Estadão’s page in GitHub. (See aditional links)

Technologies Used

Data was collected and analyzed using Python. The charts were drawn using d3.js, with final touches in Adobe Illustrator.

Project members

Rodrigo Menegat - data, infographics and story writing Bruno Ponceano - infographics Carlos Marin - web development


Additional links


Click Follow to keep up with the evolution of this project:
you will receive a notification anytime the project leader updates the project page.