Survey of Income and Program Participation (SIPP)

The primary longitudinal assessment of income fluctuation, labor force participation, social programs.

Annual tables with one record per month per person per sampled household, time period weights.
A complex sample generalizing to the U.S. civilian non-institutionalized across varying time periods.
Multi-year panels since 1980s, its current and now permanent four year rotation beginning in 2018.
Administered and financed by the US Census Bureau.

Please skim before you begin:

2023 Survey of Income and Program Participation Users’ Guide
2023 Data User Notes
A haiku regarding this microdata:

# federal programs
# poverty oversample
# monthly dynamics

Download, Import, Preparation

Determine which variables from the main table to import:

variables_to_keep <-
    c( 'ssuid' , 'pnum' , 'monthcode' , 'spanel' , 'swave' , 'erelrpe' , 
    'tlivqtr' , 'wpfinwgt' , 'rmesr' , 'thcyincpov' , 'tfcyincpov' ,
    'tehc_st' , 'rhicovann' , 'rfpov' , 'thnetworth' , 'tftotinc' )

Download and import the latest main file:

library(httr)
library(data.table)

main_tf <- tempfile()

main_url <-
    paste0(
        "https://www2.census.gov/programs-surveys/sipp/" ,
        "data/datasets/2023/pu2023_csv.zip"
    )

GET( main_url , write_disk( main_tf ) , progress() )

main_csv <- unzip( main_tf , exdir = tempdir() )

sipp_main_dt <- fread( main_csv , sep = "|" , select = toupper( variables_to_keep ) )

sipp_main_df <- data.frame( sipp_main_dt )

names( sipp_main_df ) <- tolower( names( sipp_main_df ) )

Download and import the appropriate replicate weights file:

rw_tf <- tempfile()

rw_url <-
    paste0(
        "https://www2.census.gov/programs-surveys/sipp/" ,
        "data/datasets/2023/rw2023_csv.zip"
    )

GET( rw_url , write_disk( rw_tf ) , progress() )

rw_csv <- unzip( rw_tf , exdir = tempdir() )

sipp_rw_dt <- fread( rw_csv , sep = "|" )

sipp_rw_df <- data.frame( sipp_rw_dt )

names( sipp_rw_df ) <- tolower( names( sipp_rw_df ) )

Limit both files to December records for a point-in-time estimate, then merge:

sipp_df <-
    merge(
        sipp_main_df[ sipp_main_df[ , 'monthcode' ] %in% 12 , ] ,
        sipp_rw_df[ sipp_rw_df[ , 'monthcode' ] %in% 12 , ] ,
        by = c( 'ssuid' , 'pnum' , 'monthcode' , 'spanel' , 'swave' )
    )
    
stopifnot( nrow( sipp_df ) == sum( sipp_rw_df[ , 'monthcode' ] %in% 12 ) )

Save Locally

Save the object at any point:

# sipp_fn <- file.path( path.expand( "~" ) , "SIPP" , "this_file.rds" )
# saveRDS( sipp_df , file = sipp_fn , compress = FALSE )

Load the same object:

# sipp_df <- readRDS( sipp_fn )

Survey Design Definition

Construct a complex sample survey design:

library(survey)

sipp_design <- 
    svrepdesign(
            data = sipp_df ,
            weights = ~ wpfinwgt ,
            repweights = "repwgt([1-9]+)" ,
            type = "Fay" ,
            rho = 0.5
        )

Variable Recoding

Add new columns to the data set:

rmesr_values <-
    c( 
        "With a job entire month, worked all weeks",
        "With a job all month, absent from work without pay 1+ weeks, absence not due to layoff",
        "With a job all month, absent from work without pay 1+ weeks, absence due to layoff",
        "With a job at least 1 but not all weeks, no time on layoff and no time looking for work",
        "With a job at least 1 but not all weeks, some weeks on layoff or looking for work",
        "No job all month, on layoff or looking for work all weeks",
        "No job all month, at least one but not all weeks on layoff or looking for work",
        "No job all month, no time on layoff and no time looking for work"
    )

sipp_design <- 
    
    update( 
        
        sipp_design , 
        
        one = 1 ,
        
        employment_status = factor( rmesr , levels = 1:8 , labels = rmesr_values ) ,
            
        household_below_poverty = as.numeric( thcyincpov < 1 ) ,
        
        family_below_poverty = as.numeric( tfcyincpov < 1 ) ,
        
        state_name =
            
            factor(
                
                as.numeric( tehc_st ) ,
                
                levels = 
                    c(1L, 2L, 4L, 5L, 6L, 8L, 9L, 10L, 
                    11L, 12L, 13L, 15L, 16L, 17L, 18L, 
                    19L, 20L, 21L, 22L, 23L, 24L, 25L, 
                    26L, 27L, 28L, 29L, 30L, 31L, 32L, 
                    33L, 34L, 35L, 36L, 37L, 38L, 39L, 
                    40L, 41L, 42L, 44L, 45L, 46L, 47L, 
                    48L, 49L, 50L, 51L, 53L, 54L, 55L, 
                    56L, 60L, 61L) ,
        
                labels =
                    c("Alabama", "Alaska", "Arizona", "Arkansas", "California", 
                    "Colorado", "Connecticut", "Delaware", "District of Columbia", 
                    "Florida", "Georgia", "Hawaii", "Idaho", "Illinois", "Indiana", 
                    "Iowa", "Kansas", "Kentucky", "Louisiana", "Maine", "Maryland", 
                    "Massachusetts", "Michigan", "Minnesota", "Mississippi", "Missouri", 
                    "Montana", "Nebraska", "Nevada", "New Hampshire", "New Jersey", 
                    "New Mexico", "New York", "North Carolina", "North Dakota", "Ohio", 
                    "Oklahoma", "Oregon", "Pennsylvania", "Rhode Island", "South Carolina", 
                    "South Dakota", "Tennessee", "Texas", "Utah", "Vermont", "Virginia", 
                    "Washington", "West Virginia", "Wisconsin", "Wyoming", "Puerto Rico",
                    "Foreign Country")
            )
            
    )

Analysis Examples with the `survey` library

Unweighted Counts

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

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

svyby( ~ one , ~ state_name , sipp_design , unwtd.count )

Weighted Counts

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

svytotal( ~ one , sipp_design )

svyby( ~ one , ~ state_name , sipp_design , svytotal )

Descriptive Statistics

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

svymean( ~ tftotinc , sipp_design , na.rm = TRUE )

svyby( ~ tftotinc , ~ state_name , sipp_design , svymean , na.rm = TRUE )

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

svymean( ~ employment_status , sipp_design , na.rm = TRUE )

svyby( ~ employment_status , ~ state_name , sipp_design , svymean , na.rm = TRUE )

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

svytotal( ~ tftotinc , sipp_design , na.rm = TRUE )

svyby( ~ tftotinc , ~ state_name , sipp_design , svytotal , na.rm = TRUE )

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

svytotal( ~ employment_status , sipp_design , na.rm = TRUE )

svyby( ~ employment_status , ~ state_name , sipp_design , svytotal , na.rm = TRUE )

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

svyquantile( ~ tftotinc , sipp_design , 0.5 , na.rm = TRUE )

svyby( 
    ~ tftotinc , 
    ~ state_name , 
    sipp_design , 
    svyquantile , 
    0.5 ,
    ci = TRUE , na.rm = TRUE
)

Estimate a ratio:

svyratio( 
    numerator = ~ tftotinc , 
    denominator = ~ rfpov , 
    sipp_design ,
    na.rm = TRUE
)

Subsetting

Restrict the survey design to individuals ever covered by health insurance during the year:

sub_sipp_design <- subset( sipp_design , rhicovann == 1 )

Calculate the mean (average) of this subset:

svymean( ~ tftotinc , sub_sipp_design , na.rm = TRUE )

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( ~ tftotinc , sipp_design , na.rm = TRUE )

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

grouped_result <-
    svyby( 
        ~ tftotinc , 
        ~ state_name , 
        sipp_design , 
        svymean ,
        na.rm = TRUE 
    )
    
coef( grouped_result )
SE( grouped_result )
confint( grouped_result )
cv( grouped_result )

Calculate the degrees of freedom of any survey design object:

degf( sipp_design )

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

svyvar( ~ tftotinc , sipp_design , na.rm = TRUE )

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

# SRS without replacement
svymean( ~ tftotinc , sipp_design , na.rm = TRUE , deff = TRUE )

# SRS with replacement
svymean( ~ tftotinc , sipp_design , na.rm = TRUE , deff = "replace" )

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

svyciprop( ~ family_below_poverty , sipp_design ,
    method = "likelihood" , na.rm = TRUE )

Regression Models and Tests of Association

Perform a design-based t-test:

svyttest( tftotinc ~ family_below_poverty , sipp_design )

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

svychisq( 
    ~ family_below_poverty + employment_status , 
    sipp_design 
)

Perform a survey-weighted generalized linear model:

glm_result <- 
    svyglm( 
        tftotinc ~ family_below_poverty + employment_status , 
        sipp_design 
    )

summary( glm_result )

Replication Example

This example matches statistics and standard errors from the Wealth and Asset Ownership for Households, by Type of Asset and Selected Characteristics: 2022:

Restrict the design to permanent residence-based householders to match the count in Table 4:

sipp_household_design <- subset( sipp_design , erelrpe %in% 1:2 & tlivqtr %in% 1:2 )

stopifnot( round( coef( svytotal( ~ one , sipp_household_design ) ) / 1000 , -2 ) == 134100 )

Compute Household Net Worth distribution and standard errors across the Total row of Tables 4 and 4A:

sipp_household_design <-
    update(
        sipp_household_design ,
        thnetworth_category =
            factor(
                findInterval( 
                    thnetworth , 
                    c( 1 , 5000 , 10000 , 25000 , 50000 , 100000 , 250000 , 500000 ) 
                ) ,
                levels = 0:8 ,
                labels = c( "Zero or Negative" , "$1 to $4,999" , "$5,000 to $9,999" , 
                "$10,000 to $24,999" , "$25,000 to $49,999" , "$50,000 to $99,999" , 
                "$100,000 to $249,999" , "$250,000 to $499,999" , "$500,000 or over" )
            )
    )

results <- svymean( ~ thnetworth_category , sipp_household_design )

stopifnot( 
    all.equal( as.numeric( round( coef( results ) * 100 , 1 ) ) , 
    c( 11.1 , 6.8 , 3.5 , 5.7 , 5.6 , 7.8 , 15.9 , 14.4 , 29.2 ) ) 
)

stopifnot(
    all.equal( as.numeric( round( SE( results ) * 100 , 1 ) ) ,
    c( 0.3 , 0.2 , 0.2 , 0.2 , 0.2 , 0.2 , 0.3 , 0.3 , 0.3 ) )
)

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 SIPP users, this code calculates the gini coefficient on complex sample survey data:

library(convey)
sipp_design <- convey_prep( sipp_design )

svygini( ~ tftotinc , sipp_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 SIPP users, this code replicates previously-presented examples:

library(srvyr)
sipp_srvyr_design <- as_survey( sipp_design )