This .txt file describes the data and the code required to generate the results presented in *Racial Classification as a Multistate Process*, published in Demographic Research by Jeronimo Muniz, Aliya Saperstein and Bernardo Queiroz. We conducted most statistical analysis using Stata/MP version 15.1 for Mac (64-bit Intel).



To ensure replicability, the supplemental .ZIP archive contains the following files (also available in https://github.com/blanza/race_multistate): 




1. input_ilt.xlsx - Excel file containing population and death counts structured by five-year age groups and race (White, Brown, Black)


2. input_ilt.txt - Same as above, in (1), but in text non-proprietary format


3. input_mslt.dta - Stata data file combining racial reclassification and mortality probabilities interpolated by single ages


4. input_mslt.csv - Same as above, in (3), but in comma-separated values format


5. replication_code.do - Stata code (for Stata versions 10.0 and beyond) to replicate tables and figure in the article. Note that manual editing is needed to generate the final tables in the article. Also note that two Stata packages are required to estimate race-specific single decrement life tables (ilt) and multistate life tables (mslt). More about these Stata programs are in: 



	Muniz, Jeronimo O. (2020) Multistate life tables using Stata. The Stata Journal 20.3: 721-745.

	Muniz, Jeronimo O. (2023) Iterative intercensal single-decrement life tables using Stata. The Stata Journal 23.3: 813-834.





The final working files enumerated above were generated using raw baseline data from two official statistical agencies in Brazil (IBGE and SIM, Datasus, Ministério da Saude 2021). These baseline data were retrieved from online sources before December 2023.  



PUBLICLY AVAILABLE DATA USED TO PRODUCE FILES in (1) AND (2):



a. Population counts by age and race:

	i. 2010 Population Census (www.ibge.gov.br). Also available as additional files from Miranda (2015): https://doi.org/10.4054/DemRes.2015.32.59

	ii. 2019 Population estimated from the Pesquisa Nacional por Amostra de Domicilios – Continua (PNAD-C) (https://sidra.ibge.gov.br/tabela/6403)

b. Mortality by age and race (death counts):

	i. Ministry of Health’s Mortality Information System (SIM, Datasus, Ministério da Saúde 2021 (https://opendatasus.saude.gov.br/dataset/sim)

PUBLICLY AVAILABLE DATA USED TO PRODUCE FILES in (3) AND (4):



c. Probabilities of racial reclassification were calculated using data from the Panel 6 of the 2017-2019 PNAD-C:

	i. PNAD-C: https://ftp.ibge.gov.br/Trabalho_e_Rendimento/Pesquisa_Nacional_por_Amostra_de_Domicilios_continua/Trimestral/Microdados


	ii. Data were organized and extracted using the code and individual linkage algorithms (Ribas and Soares 2009) available in Data Zoom: https://www.econ.puc-rio.br/datazoom/english/pnadc.html