Paper on ChatGPT Political Bias
More Human than Human: Measuring ChatGPT Political Bias.
We investigate the political bias of a large language model (LLM), ChatGPT, which has become popular for retrieving factual information and generating content. Although ChatGPT assures that it is impartial, the literature suggests that LLMs exhibit bias involving race, gender, religion, and political orientation. Political bias in LLMs can have adverse political and electoral consequences similar to bias from traditional and social media. Moreover, political bias can be harder to detect and eradicate than gender or racial bias. We propose a novel empirical design to infer whether ChatGPT has political biases by requesting it to impersonate someone from a given side of the political spectrum and comparing these answers with its default. We also propose dose-response, placebo, and profession-politics alignment robustness tests. To reduce concerns about the randomness of the generated text, we collect answers to the same questions 100 times, with question order randomized on each round. We find robust evidence that ChatGPT presents a significant and systematic political bias toward the Democrats in the US, Lula in Brazil, and the Labour Party in the UK. These results translate into real concerns that ChatGPT, and LLMs in general, can extend or even amplify the existing challenges involving political processes posed by the Internet and social media. Our findings have important implications for policymakers, media, politics, and academia stakeholders.
Covered by almost 170 pieces of news coverage within the first five days after publication – including the Washington Post, Telegraph, Daily Mail, Sky News, Forbes, Yahoo News Radio 5 Live, and many more national and international outlets. See the attached list.
Paper on Inequalities in access to delivery services in Brazil
Inequalities in the geographic access to delivery services in Brazil
We observe that, on average, (i) the share of women traveling for childbirth increased, reaching 31% in 2017, and (ii) distances got longer, approaching the 60-kilometer mark by 2017. The increase in distance is mainly due to more women traveling. Nevertheless, regional disparities persist, especially between the north/inland and coastal regions. Women with high-risk pregnancies or newborns with risks such as low birth weight tend to travel longer distances. However, those residing in higher-development municipalities tend to travel shorter distances.