Compounding Effects iPUMS Income Pull

Household examples to pull income for:

Single Mother + Child who receive SNAP + Medicaid
2 Parents + Child + Elderly
Single Parent + Child + Elderly
Pregnant + Child + Elderly

Adding ipums data

FOODSTMP = food stamp recipient?
- 0 = n/a
- 1 = no
- 2 = yes
AGE by single year
POVERTY (we want <=100)
- 000 = NA
- 001 = 1 percent or less of poverty threshold (including 0 or negative income)
- 501 = 501 percent or more of poverty threshold
HINSCAID = medicaid status (can use as proxy for snap eligibility?)
- 1 = no insurance through medicaid
- 2 = yes insurance through medicaid
INCWELFR = income from welfare programs
- 99999 = N/A
HHINCOME = Total household income
- 9999999 = N/A
MULTGEN = Multigenerational HH
- 2 gen = 2
FAMSIZE = family size

Modifications of Kaitlan’s Code

library(ipumsr)
library(dplyr)


Attaching package: 'dplyr'

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

# Note to self: Must set wd to where all files are, then open ddi and feed into read_ipums_micro
ddi <- read_ipums_ddi("usa_00074.xml")
data_raw <- read_ipums_micro(ddi)

Use of data from IPUMS USA is subject to conditions including that users should cite the data appropriately. Use command `ipums_conditions()` for more details.

data <- data_raw
names(data)

 [1] "YEAR"       "MULTYEAR"   "SAMPLE"     "SERIAL"     "CBSERIAL"  
 [6] "HHWT"       "CLUSTER"    "STATEFIP"   "STRATA"     "GQ"        
[11] "HHINCOME"   "FOODSTMP"   "COUPLETYPE" "MULTGEN"    "MULTGEND"  
[16] "PERNUM"     "PERWT"      "FAMUNIT"    "FAMSIZE"    "NCHILD"    
[21] "ELDCH"      "YNGCH"      "RELATE"     "RELATED"    "SEX"       
[26] "AGE"        "HINSCAID"   "INCWELFR"   "INCSUPP"    "POVERTY"

data$inc_c <- ifelse(as.numeric(data$HHINCOME)<9999999,as.numeric(data$HHINCOME),NA)
data$pov_c <- ifelse(as.numeric(data$POVERTY)>1 & as.numeric(data$POVERTY)<201 ,1,0)

#create person-level flags (stealing from Kaitlan's code)
data <- data %>%
  mutate(
    is_head = RELATE == 1,
    is_spouse = RELATE == 2,
    is_elderly = AGE >= 65,
    is_child = AGE < 18,
    is_pov = pov_c==1,
    is_female_head = is_head & SEX == 2,
    has_medicaid_person = HINSCAID == 2,
    has_foodstamps_person = FOODSTMP == 2,
    child_on_medicaid = is_child & has_medicaid_person,
    elderly_on_medicaid = is_elderly & has_medicaid_person
  )


#collapse to household-level
hh_summary <- data %>%
  group_by(SERIAL) %>%
  summarise(
    hh_income = first(inc_c),
    hh_weight = first(HHWT),
    fam_sz = first(FAMSIZE),
    num_pov = sum(is_pov),
    has_children = any(is_child),
    has_elderly = any(is_elderly),
    num_medicaid = sum(has_medicaid_person),
    num_parents = sum(is_head | is_spouse),
    num_children = sum(is_child),
    num_elderly = sum(is_elderly),
    num_foodstamps = sum(has_foodstamps_person),
    num_children_medicaid = sum(child_on_medicaid),
    num_elderly_medicaid = sum(elderly_on_medicaid),
    female_head = any(is_female_head),
    any_foodstamps = any(has_foodstamps_person,na.rm=FALSE),
    any_children_medicaid = all(child_on_medicaid, na.rm=FALSE),
    any_elderly_medicaid = all(elderly_on_medicaid, na.rm=FALSE)
  )



#filter for households with children and elderly on Medicaid and all hh members on Food Stamps
hh_summary <- hh_summary %>%
  filter(num_foodstamps>0 & num_medicaid>0 & num_pov>0)

#classify household types
hh_summary <- hh_summary %>%
mutate(
  household_type = case_when(
    num_parents == 1 & female_head==T & num_children == 1 & !has_elderly & fam_sz==2 &num_children_medicaid>0  ~ "Single Mother + Child",
    num_parents == 1 & num_children == 1 & fam_sz==2 & !has_elderly & num_children_medicaid>0 ~ "Single Parent + Child",
    num_parents == 2 & num_children == 1 & has_elderly & num_elderly_medicaid>0 ~ "Two Parents + Child + Elderly",
    num_parents == 2 & num_children == 1 & fam_sz==3 & !has_elderly & num_children_medicaid>0 ~ "Two Parents + Child",
    TRUE ~ "Other"
  )
)


#calculate weighted median income by household type
#install.packages("matrixStats")
library(matrixStats)


Attaching package: 'matrixStats'

The following object is masked from 'package:dplyr':

    count

weighted_median_income <- hh_summary %>%
  group_by(household_type) %>%
  summarise(n=n(),
            tothh=sum(hh_weight),
    weighted_median_income = weightedMedian(hh_income, w = hh_weight, na.rm = TRUE)
  )

print(weighted_median_income)

# A tibble: 5 × 4
  household_type                    n  tothh weighted_median_income
  <chr>                         <int>  <dbl>                  <dbl>
1 Other                         25306 636063                 22964 
2 Single Mother + Child          1431  42774                 16322 
3 Single Parent + Child           139   3734                 22674.
4 Two Parents + Child             603  15936                 25280.
5 Two Parents + Child + Elderly   123   2488                 36352.