Application of the Network Scale-Up Method to Estimate Hard-to-Count Populations: The Case of Brazilian Immigrants in Florida, USA

Dimitri Fazito, Centro de Desenvolvimento e Planejamento Regional (CEDEPLAR)
Chris McCarty, University of Florida

According to the US Census Bureau there were 212,428 Brazilian immigrants in the country in 2000 but concurrently the Brazilian government (extra-officially), the media and varied researchers argued that this is an underestimate and claimed different numbers ranging from 400,000 to 1,500,000. In this study we propose to estimate the subpopulation of Brazilian immigrant residents in Florida, USA (including estimates about Mexican immigrants) based on the network scale-up (NSU) method. This method aims to estimate the size of (hard to count) unknown subpopulations assuming that the fraction of these subpopulations presented in the averaged active social networks of the general population members could be scaled up to their real sizes. In this paper we present the NSU rationale, explain its general application to a case study and conclude with remarks about the use of NSU in demography.