Geographical comparability of the quality of data in the Brazilian censuses?


I am working on the 1980, 1991, 2000 and 2010 Brazilian censuses.

Brazil is continent-sized country so I am wondering whether the quality of the micro-data is the same in every region, state or municipality.

For instance, the whole Northern region was excluded from censuses prior to 1980 so the IBGE (the Brazilian institute organizing the censuses) may have less experience gathering data there. Relatedly I guess, the Northern region is almost entirely made of the Amazonian forest so I guess it is harder to collect data there.

I have looked for some academic discussion about this specific topic but have not found any.

Do you have any idea, or at least any reference about whether the data quality of the Censuses is debatable in some parts of Brazil?

Thank you for your help.


I am afraid I have not found any official statements speaking to your specific concerns. While the IBGE have gone to great lengths to ensure the validity of the data, your assumptions about the difficulty of gathering census information in the norther regions seem reasonable. You may be able to test your hypothesis by checking the variability of information in that region from 1980 and 1991 vs 2000 and 2010.

I am sorry I could not offer a more satisfying answer. If you haven’t already, I would also recommend trying to get in touch with IBGE directly.