Pesquisa de Orcamentos Familiares (POF)

Brazil’s household budget survey designed to guide major economic indicators like the Contas nacionais.

Various tables with one record per sampled household, resident, job, expenditure.
A complex sample survey designed to generalize to the civilian population of Brazil.
Released at irregular intervals, 2002-2003, 2008-2009, and 2017-2018 microdata available.
Administered by the Instituto Brasileiro de Geografia e Estatistica.

Please skim before you begin:

Pesquisa de Orçamentos Familiares 2017-2018 Perfil das despesas no Brasil
Conceitos e métodos
This human-composed haiku or a bouquet of artificial intelligence-generated limericks

# shopping na praia
# roupa, comida, pede
# tres havaianas

Download, Import, Preparation

Download the dictionary files:

library(archive)

dictionary_tf <- tempfile()

dictionary_url <-
    paste0(
        "https://ftp.ibge.gov.br/Orcamentos_Familiares/" ,
        "Pesquisa_de_Orcamentos_Familiares_2017_2018/Microdados/Documentacao_20230713.zip"
    )

download.file( dictionary_url , dictionary_tf , mode = 'wb' )

dictionary_files <- archive_extract( dictionary_tf , dir = tempdir() )

Import the household variable dictionary:

library(readxl)

dictionary_fn <- file.path( tempdir() , "Dicionários de váriaveis.xls" )

domicilio_dictionary_tbl <- read_excel( dictionary_fn , sheet = "Domicílio" , skip = 3 )

domicilio_dictionary_df <- data.frame( domicilio_dictionary_tbl )

names( domicilio_dictionary_df ) <-
    c( 'position' , 'length' , 'decimals' , 'column_name' , 'description' , 'variable_labels' )

domicilio_dictionary_df[ c( 'position' , 'length' , 'decimals' ) ] <-
    sapply( domicilio_dictionary_df[ c( 'position' , 'length' , 'decimals' ) ] , as.integer )

domicilio_dictionary_df <- subset( domicilio_dictionary_df , !is.na( position ) )

Import the resident variable dictionary:

morador_dictionary_tbl <- read_excel( dictionary_fn , sheet = "Morador" , skip = 3 )

morador_dictionary_df <- data.frame( morador_dictionary_tbl )

names( morador_dictionary_df ) <-
    c( 'position' , 'length' , 'decimals' , 'column_name' , 'description' , 'variable_labels' )

morador_dictionary_df[ c( 'position' , 'length' , 'decimals' ) ] <-
    sapply( morador_dictionary_df[ c( 'position' , 'length' , 'decimals' ) ] , as.integer )

morador_dictionary_df <- subset( morador_dictionary_df , !is.na( position ) )

Import the post-stratification totals:

post_stratification_fn <- file.path( tempdir() , "Pos_estratos_totais.xlsx" )

post_stratification_tbl <- read_excel( post_stratification_fn , skip = 5 )
    
post_stratification_df <- data.frame( post_stratification_tbl )

names( post_stratification_df ) <-
    c( 'estrato_pof' , 'pos_estrato' , 'total_pessoas' , 'uf' , 'cod_upa' )

Download the full dataset:

this_tf <- tempfile()

this_url <-
    paste0(
        "https://ftp.ibge.gov.br/Orcamentos_Familiares/" ,
        "Pesquisa_de_Orcamentos_Familiares_2017_2018/Microdados/Dados_20230713.zip"
    )

download.file( this_url , this_tf , mode = 'wb' )

unzipped_files <- unzip( this_tf , exdir = tempdir() )

Import the household table:

library(readr)

domicilio_fn <- grep( 'DOMICILIO\\.txt$' , unzipped_files , value = TRUE )

domicilio_tbl <-
    read_fwf(
        domicilio_fn ,
        fwf_widths( 
            widths = domicilio_dictionary_df[ , 'length' ] , 
            col_names = domicilio_dictionary_df[ , 'column_name' ] 
        )
    )

domicilio_df <- data.frame( domicilio_tbl )

names( domicilio_df ) <- tolower( names( domicilio_df ) )

Import the resident table:

morador_fn <- grep( 'MORADOR\\.txt$' , unzipped_files , value = TRUE )

morador_tbl <-
    read_fwf(
        morador_fn ,
        fwf_widths( 
            widths = morador_dictionary_df[ , 'length' ] , 
            col_names = morador_dictionary_df[ , 'column_name' ] 
        )
    )

morador_df <- data.frame( morador_tbl )

names( morador_df ) <- tolower( names( morador_df ) )

Merge one household-level variable and also the post-stratification info onto the person-level table:

dom_mor_df <- merge( domicilio_df[ c( 'cod_upa' , 'num_dom' , 'v6199' ) ] , morador_df )

pof_df <- merge( dom_mor_df , post_stratification_df )

stopifnot( nrow( pof_df ) == nrow( morador_df ) )

Save locally

Save the object at any point:

# pof_fn <- file.path( path.expand( "~" ) , "POF" , "this_file.rds" )
# saveRDS( pof_df , file = pof_fn , compress = FALSE )

Load the same object:

# pof_df <- readRDS( pof_fn )

Survey Design Definition

Construct a complex sample survey design:

library(survey)

options( survey.lonely.psu = "adjust" )

pre_stratified_design <- 
    svydesign(
        id = ~ cod_upa , 
        strata = ~ estrato_pof ,
        weights = ~ peso ,
        data = pof_df ,
        nest = TRUE
    )

population_totals <- 
    aggregate( peso_final ~ pos_estrato , data = pof_df , sum )
    
names( population_totals ) <- c( 'pos_estrato' , 'Freq' )

pof_design <-
    postStratify(
        pre_stratified_design , 
        ~ pos_estrato , 
        population_totals
    )

Variable Recoding

Add new columns to the data set:

pof_design <-
    update(
        pof_design ,
        
        one = 1 ,
        
        food_security =
            factor( 
                v6199 , 
                levels = 1:4 , 
                labels = c( 'food secure' , 'mild' , 'moderate' , 'severe' ) 
            ) ,
    
        age_categories =
            factor( 
                1 + findInterval( v0403 , 
                    c( 20 , 25 , 30 , 35 , 45 , 55 , 65 , 75 ) ) ,
                levels = 1:9 , 
                labels =
                    c( "under 20" , "20-24" , "25-29" , "30-34" , "35-44" , 
                    "45-54" , "55-64" , "65-74" , "75+" )
            ) ,
        
        sex = factor( v0404 , levels = 1:2 , labels = c( 'male' , 'female' ) ) ,
        
        urban = as.numeric( tipo_situacao_reg == 1 )

    )

Analysis Examples with the `survey` library

Unweighted Counts

Count the unweighted number of records in the survey sample, overall and by groups:

sum( weights( pof_design , "sampling" ) != 0 )

svyby( ~ one , ~ sex , pof_design , unwtd.count )

Weighted Counts

Count the weighted size of the generalizable population, overall and by groups:

svytotal( ~ one , pof_design )

svyby( ~ one , ~ sex , pof_design , svytotal )

Descriptive Statistics

Calculate the mean (average) of a linear variable, overall and by groups:

svymean( ~ renda_total , pof_design )

svyby( ~ renda_total , ~ sex , pof_design , svymean )

Calculate the distribution of a categorical variable, overall and by groups:

svymean( ~ age_categories , pof_design )

svyby( ~ age_categories , ~ sex , pof_design , svymean )

Calculate the sum of a linear variable, overall and by groups:

svytotal( ~ renda_total , pof_design )

svyby( ~ renda_total , ~ sex , pof_design , svytotal )

Calculate the weighted sum of a categorical variable, overall and by groups:

svytotal( ~ age_categories , pof_design )

svyby( ~ age_categories , ~ sex , pof_design , svytotal )

Calculate the median (50th percentile) of a linear variable, overall and by groups:

svyquantile( ~ renda_total , pof_design , 0.5 )

svyby( 
    ~ renda_total , 
    ~ sex , 
    pof_design , 
    svyquantile , 
    0.5 ,
    ci = TRUE 
)

Estimate a ratio:

svyratio( 
    numerator = ~ renda_total , 
    denominator = ~ anos_estudo , 
    pof_design ,
    na.rm = TRUE
)

Subsetting

Restrict the survey design to credit card holders:

sub_pof_design <- subset( pof_design , v0409 > 0 )

Calculate the mean (average) of this subset:

svymean( ~ renda_total , sub_pof_design )

Measures of Uncertainty

Extract the coefficient, standard error, confidence interval, and coefficient of variation from any descriptive statistics function result, overall and by groups:

this_result <- svymean( ~ renda_total , pof_design )

coef( this_result )
SE( this_result )
confint( this_result )
cv( this_result )

grouped_result <-
    svyby( 
        ~ renda_total , 
        ~ sex , 
        pof_design , 
        svymean 
    )
    
coef( grouped_result )
SE( grouped_result )
confint( grouped_result )
cv( grouped_result )

Calculate the degrees of freedom of any survey design object:

degf( pof_design )

Calculate the complex sample survey-adjusted variance of any statistic:

svyvar( ~ renda_total , pof_design )

Include the complex sample design effect in the result for a specific statistic:

# SRS without replacement
svymean( ~ renda_total , pof_design , deff = TRUE )

# SRS with replacement
svymean( ~ renda_total , pof_design , deff = "replace" )

Compute confidence intervals for proportions using methods that may be more accurate near 0 and 1. See ?svyciprop for alternatives:

svyciprop( ~ urban , pof_design ,
    method = "likelihood" )

Regression Models and Tests of Association

Perform a design-based t-test:

svyttest( renda_total ~ urban , pof_design )

Perform a chi-squared test of association for survey data:

svychisq( 
    ~ urban + age_categories , 
    pof_design 
)

Perform a survey-weighted generalized linear model:

glm_result <- 
    svyglm( 
        renda_total ~ urban + age_categories , 
        pof_design 
    )

summary( glm_result )

Replication Example

This example matches the 2017-2018 person-level food security estimates from Tabela 3:

person_level_food_security <- svymean( ~ food_security , pof_design , na.rm = TRUE )
    
stopifnot(
    all.equal(
        round( coef( person_level_food_security ) , 2 ) , 
        c( 0.59 , 0.27 , 0.09 , 0.05 ) , 
        check.attributes = FALSE 
    )
)

Poverty and Inequality Estimation with `convey`

The R convey library estimates measures of income concentration, poverty, inequality, and wellbeing. This textbook details the available features. As a starting point for POF users, this code calculates the gini coefficient on complex sample survey data:

library(convey)
pof_design <- convey_prep( pof_design )

svygini( ~ renda_total , pof_design , na.rm = TRUE )

Analysis Examples with `srvyr`

The R srvyr library calculates summary statistics from survey data, such as the mean, total or quantile using dplyr-like syntax. srvyr allows for the use of many verbs, such as summarize, group_by, and mutate, the convenience of pipe-able functions, the tidyverse style of non-standard evaluation and more consistent return types than the survey package. This vignette details the available features. As a starting point for POF users, this code replicates previously-presented examples:

library(srvyr)
pof_srvyr_design <- as_survey( pof_design )