Household Type Estimates (under 200% poverty threshold)

Estimates of households for the different scenarios used in the compounding effects analysis.

For this second draft, instead of using income cutoffs, I filter by poverty status (households under 200% poverty line).

Data extract was filtered to Texas before download.

There are no ACS variables to measure whether a person is pregnant. Alternatively, I included scenarios for single mother households.

library(ipumsr)
library(dplyr)
library(writexl)

ddi <- read_ipums_ddi("usa_00051.xml")
ipums_data <- read_ipums_micro(ddi, data_file = "usa_00051.dat", verbose = FALSE)

# remove NAs
ipums_clean <- ipums_data %>%
  filter(
    !is.na(AGE),
    !is.na(SEX),
    !is.na(RELATE),
    !is.na(HHINCOME),
    !is.na(HHWT),
    !is.na(SERIAL),
    !is.na(POVERTY)
  )

# define roles and flags
hh_data <- ipums_clean %>%
  mutate(
    is_child = RELATE == 3 & AGE < 18,
    is_elderly = AGE >= 65,
    is_female = SEX == 2,
    is_householder = RELATE == 1,
    is_spouse = RELATE == 2
  )

hh_summary <- hh_data %>%
  group_by(SERIAL) %>%
  summarise(
    n_children = sum(is_child),
    n_elderly = sum(is_elderly),
    has_female_householder = any(is_householder & is_female),
    n_parents = sum(RELATE %in% c(1, 2)),
    hh_income = max(HHINCOME),
    hh_weight = max(HHWT),
    poverty = first(POVERTY),
    .groups = "drop"
  )

# filter for households under 200% poverty
hh_summary <- hh_summary %>%
  filter(poverty < 200)

# Scenario 1: Female parent + 1 child + income ≤ $19,307
scenario1 <- hh_summary %>%
  filter(
    has_female_householder,
    n_parents == 1,
    n_children == 1,
    n_elderly == 0
  )

# Scenario 2: Single parent (any gender) + 1 child + 1 elderly + income ≤ $19,307
scenario2 <- hh_summary %>%
  filter(
    n_parents == 1,
    n_children == 1,
    n_elderly == 1
  )

# Scenario 3: Two parents + 1 child + 1 elderly + income ≤ $25,678
scenario3 <- hh_summary %>%
  filter(
    n_parents == 2,
    n_children == 1,
    n_elderly == 1
  )

# Scenario 4: Female parent + 1 child + 1 elderly + income ≤ $19,307
scenario4 <- hh_summary %>%
  filter(
    has_female_householder,
    n_parents == 1,
    n_children == 1,
    n_elderly == 1
  )

# Weighted household counts
scenario_counts <- list(
  Scenario1 = sum(scenario1$hh_weight),
  Scenario2 = sum(scenario2$hh_weight),
  Scenario3 = sum(scenario3$hh_weight),
  Scenario4 = sum(scenario4$hh_weight)
)

print(scenario_counts)

$Scenario1
[1] 158966

$Scenario2
[1] 7096

$Scenario3
[1] 7217

$Scenario4
[1] 4874

#save to Excel
write_xlsx(as.data.frame(scenario_counts), "hh_pov.xlsx")