Designing Inclusive National Surveys: A Participatory and Adaptive Approach to Sampling LGBTQIA+ Populations in Brazil

Samuel da Silva , Cidacs/Fiocruz-BA
Fernanda Fortes de Lena, Centre d' Estudis Demogràfics
Ana Hermeto, UFMG - Cedeplar
Raissa Sidrim, Instituto Matizes
Juliana Oliveira, Instituto Matizes
Lucas Bulgarelli , Instituto Matizes
Arthur Fontgaland, Instituto Matizes
Ju Motter, Instituto Matizes
Hannah Maruci, Instituto Matizes
Mariah Rafaela da Silva, World Bank

This paper discusses the development of a participatory and adaptive sampling design employed in the Economic Cost of Exclusion in Brazil Survey (ECOE 2025), a national study investigating labor market exclusion among LGBTQIA+ populations. It addresses the research question: What steps are required to create a more LGBTQIA+ inclusive sample in surveys that focus on labor market outcomes? The survey was conceived to fill a critical gap in Brazil’s data infrastructure. While the National Health Survey (PNS) includes questions on sexual orientation, it exhibits high non-response bias and lacks labor market indicators. Conversely, the Continuous National Household Sample Survey (PNAD Contínua) provides detailed labor data but does not identify LGBTQIA+ individuals. The ECOE 2025 bridges this disconnect by integrating visibility and socioeconomic dimensions within an inclusive, adaptive, and community-based framework. The online survey reached 11,206 valid respondents across all Brazilian regions through multiple recruitment strategies, including digital campaigns, partnerships with community organizations, and in-person outreach in shelters, cultural centers, and peripheral urban areas. Continuous benchmarking against the PNAD Contínua (2023) and the 2022 Census guided real-time adjustments to improve inclusion of underrepresented groups such as trans and travesti persons, quilombola and Indigenous populations, and residents of informal settlements. Sixteen qualitative focus groups contextualized emerging quantitative patterns. The resulting sample demonstrates broad demographic and territorial diversity and greater representativity across gender, education, and regional dimensions than official surveys, illustrating the potential of participatory and adaptive methods to reduce compositional bias in national labor market research.

See extended abstract

 Presented in Session P8. Demographic Trends, History, Data and Methods