# Food Security Fridays — Week 13: Race and Ethnicity Reproducibility Package

## "The Groups You Won't Find in the Annual Report"

**Author:** Matthew P. Rabbitt, PhD
**Date:** April 2026
**Series:** Food Security Fridays

---

## Overview

This post disaggregates U.S. food insecurity rates by race and ethnicity
beyond the four categories published in the USDA annual food security
report (ERR-358): White non-Hispanic, Black non-Hispanic, Hispanic, and
Other non-Hispanic. Using the detailed race (PTDTRACE), detailed Hispanic
origin (PRDTHSP), and detailed Asian origin (PRDASIAN) recodes on the
December 2024 CPS Food Security Supplement, the analysis produces
disaggregated food insecurity and SNAP take-up rates inside the "Other
non-Hispanic", "Hispanic", and "Asian alone" categories.

> **Context: CPS-FSS cancellation.** USDA cancelled the CPS Food Security
> Supplement in September 2025. The December 2024 collection analyzed
> here is the **final CPS-FSS**. The disaggregation this package produces
> is therefore not a "preview of things to come" — it is the definitive
> set of detailed race/ethnicity food insecurity estimates available from
> the CPS-FSS series. Any future food insecurity data collection that
> replaces the CPS-FSS will need to be benchmarked against these final
> ERR-358 numbers, and any MENA food insecurity statistic will depend on
> a successor survey adopting the 2024 SPD 15 combined question.

The exercise is motivated by the March 2024 revision to OMB Statistical
Policy Directive No. 15 (SPD 15), which (a) replaces the two-question
race/ethnicity format with a single combined question, (b) adds Middle
Eastern or North African (MENA) as a minimum reporting category, and
(c) requires agencies to publish disaggregated estimates within minimum
categories when sample sizes permit. Federal agencies must comply by
March 2029. Because the CPS-FSS was cancelled before implementing the
new standard, the detailed Asian, detailed Hispanic, and AIAN/NHPI/
Multiracial estimates produced here are also the last CPS-FSS rates that
will ever be published under the 1997 standard.

## Unit of Analysis: Household

**Every rate in this post and infographic is a household-level rate — the
share of *households* (not persons) in a given race/ethnicity group that
are food insecure.** The reproducibility file:

1. Loads the December 2024 CPS-FSS supplement-respondent file
   (individuals in households that received the food security supplement).
2. Keeps one record per household, preferring the CPS reference person
   (PERRP = 40 with relatives, or PERRP = 41 living alone). This yields
   N = 32,655 households with valid food security status.
3. Classifies each household by the reference person's race (PTDTRACE),
   Hispanic origin (PEHSPNON, PRDTHSP), and — for Asian-alone households
   — detailed Asian origin (PRDASIAN).
4. Weights estimates by the household supplement weight (HHSUPWGT),
   which Census stores with four implied decimals.
5. Classifies food security status from the pre-constructed HRFS12M1
   variable (1 = High/Marginal, 2 = Low, 3 = Very Low), which is itself
   defined at the household level from the 18-item HFSSM.

This matches the USDA ERS convention used in ERR-358. Aggregate weighted
totals (134,062K households, 18,341K food-insecure, 13.7% FI rate)
reproduce **ERR-358 Table 1A** (the all-households line) exactly, and the
four-category race/ethnicity FI rates in Table 1 of this package reproduce
**ERR-358 Table 2** (panel labelled "Race and ethnicity of households") to
within rounding:

| Group | This file | ERR-358 Table 2 |
|-------|----------:|----------------:|
| White, non-Hispanic | 10.1% | 10.1% |
| Black, non-Hispanic | 24.4% | 24.4% |
| Hispanic (any race) | 20.2% | 20.2% |
| Other, non-Hispanic | 11.9% | 11.9% |

**Important caveat.** Because race/Hispanic origin is assigned from the
reference person, multi-race or multi-ethnic households are counted
entirely under the reference person's identity. A household with a
White non-Hispanic reference person and Asian or Hispanic children, for
example, is classified as White non-Hispanic. This is the same
convention ERR-358 uses, so our disaggregated estimates are directly
comparable to the annual report. A person-level analysis (assigning
each individual their own race/ethnicity and weighting by the person
supplement weight) would produce different denominators and slightly
different rates, particularly for groups with large differences between
household- and person-level composition (e.g., Hispanic and Asian
households, which are larger on average than White non-Hispanic
households).

## Contents

| File | Description |
|------|-------------|
| `fsf_week13_reproducibility.do` | Self-contained Stata analysis (11 tables) |
| `cpsdec2024.do` | CPS-FSS data cleaning program |
| `fsf_week13_README.md` | This file |
| `fsf_week13_additional_statistics.md` | Supplementary statistics: SPD 15 details, VLFS, child FI, income-controlled FI, multiracial detail, geographic concentration, prior ERS acknowledgment |

## Data Requirements

- **December 2024 CPS Food Security Supplement** raw data file (`dec24pub.dat`)
  — the **final CPS-FSS** before USDA cancelled the series in September 2025.
- Source: U.S. Census Bureau, Current Population Survey
- Available at: https://www.census.gov/data/datasets/time-series/demo/cps/cps-supp_cps-repwgt/cps-food-security.html

## Instructions

1. Download the December 2024 CPS-FSS raw ASCII file (`dec24pub.dat`) from
   the Census Bureau URL above.
2. Open `cpsdec2024.do` and update the two `local` path macros at the top
   (`indir` and `outdir`) to point to (a) the directory containing
   `dec24pub.dat` and (b) the directory where `cpsdec2024.dta` should be
   saved. Then run `cpsdec2024.do` to produce `cpsdec2024.dta`.
3. Open `fsf_week13_reproducibility.do` and update the `$rawdata` and
   `$projdir` globals (at the top of the file) to point to (a) the
   directory containing `cpsdec2024.dta` and (b) the project working
   directory for output and logs.
4. Run `fsf_week13_reproducibility.do`.

The analysis log and any output files will be saved to `$projdir/output/`.

## Analysis Tables

| Table | Description |
|-------|-------------|
| 1 | ERR-358 4-category reproduction (White NH, Black NH, Hispanic, Other NH) |
| 2 | Decomposition of "Other non-Hispanic" (AIAN, Asian, NHPI, Multiracial) |
| 3 | Within-Hispanic disaggregation by detailed origin (8 groups) |
| 4 | Within-Asian disaggregation by detailed origin (7 groups) |
| 5 | Multiracial non-Hispanic households by race combination |
| 6 | Very-low-food-security rates by detailed group |
| 7 | SNAP participation among food-insecure households by detailed group |
| 8 | Summary statistics for infographic |
| 9 | Design-effect-adjusted 95% confidence intervals (deff = 1.5) for every detailed group |
| 10 | Child-referenced food insecurity (HRFS12MC) by detailed race/ethnicity |
| 11 | Food insecurity above 185% FPL (income-to-poverty ratio ≥ 1.85 from midpoint-imputed income and 2024 Census poverty thresholds) — income-controlled disaggregation |

## Key Variables

### Race/ethnicity assignment

Race and Hispanic origin are assigned at the household level from the
CPS reference person (PERRP = 40 or 41). The reproducibility file
constructs the following hierarchical, mutually exclusive typology:

| Variable | Definition |
|----------|-----------|
| `raceeth4` | ERR-358 4-category (White NH, Black NH, Hispanic, Other NH) |
| `raceeth8` | 7-category disaggregation (White NH, Black NH, AIAN NH, Asian NH, NHPI NH, Multiracial NH, Hispanic any race) |
| `hisp_detail` | 8-category within-Hispanic (Mexican, Puerto Rican, Cuban, Dominican, Salvadoran, Central American excl. Salvadoran, South American, Other Spanish) |
| `asian_detail` | 7-category within-Asian NH (Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, Other Asian) |
| `multi_group` | Multiracial NH combinations (White-Black, White-AIAN, White-Asian, White-NHPI, Black-AIAN/Asian/NHPI, Other two-race, Three or more races) |
| `hhincome` | Household income, midpoint-imputed from `HEFAMINC` brackets (top bracket assigned $160K) |
| `hpvcut` | 2024 Census poverty threshold assigned by `numpers`, `numch`, and reference-person age |
| `hhincpv` | Household income-to-poverty ratio (`hhincome / hpvcut`); used for the Table 11 above-185% FPL restriction |

### CPS source variables

| Variable | Description |
|----------|-------------|
| `PTDTRACE` | 26-category detailed race recode |
| `PEHSPNON` | Hispanic non-Hispanic indicator |
| `PRDTHSP` | 8-category detailed Hispanic origin |
| `PRDASIAN` | 7-category detailed Asian subgroup (populated when PTDTRACE indicates Asian) |
| `PERRP` | Relationship to reference person (40 = reference with relatives, 41 = reference alone) |
| `HRFS12M1` | 3-category household food security status (pre-constructed) |
| `HRFS12MC` | Children's food security status (pre-constructed) |
| `HHSUPWGT` | Household supplement weight (4 implied decimals) |
| `HESP1`/`HESP2` | SNAP receipt indicators |
| `HEFAMINC` | Family income bracket (16 categories, used for midpoint imputation) |
| `HXFAMINC` | Family-income allocation flag |
| `HRNUMHOU` | Number of persons in household |
| `PRTAGE` | Person's age (used for reference-person age in poverty threshold) |

## Key Statistics

| Finding | Value | Source Table |
|---------|-------|-------------|
| Overall food insecurity rate | 13.7% | Table 1 |
| Food-insecure households | 18.34M | Table 1 |
| Total households | 134.06M | Table 1 |
| White NH FI rate | 10.1% | Table 1 |
| Black NH FI rate | 24.4% | Table 1 |
| Hispanic FI rate | 20.2% | Table 1 |
| Other NH FI rate | 11.9% | Table 1 |
| AIAN alone NH FI rate | 30.9% (N=272) | Table 2 |
| AIAN alone NH VLFS rate | 14.5% | Table 2 |
| Asian alone NH FI rate | 6.6% | Table 2 |
| NHPI alone NH FI rate | 18.4% (N=108) | Table 2 |
| Multiracial NH FI rate | 21.8% | Table 2 |
| Range inside "Other NH" | 4.7x (30.9% vs 6.6%) | Table 2 |
| Puerto Rican FI rate | 24.9% | Table 3 |
| Cuban FI rate | 17.3% | Table 3 |
| Range inside Hispanic | 7.6 pp (24.9% vs 17.3%) | Table 3 |
| Vietnamese FI rate | 12.5% (N=129) | Table 4 |
| Asian Indian FI rate | 3.2% | Table 4 |
| Range inside Asian NH | 3.9x (12.5% vs 3.2%) | Table 4 |
| AIAN SNAP (if FI) | 52.1% | Table 7 |
| Puerto Rican SNAP (if FI) | 51.4% | Table 7 |
| South American SNAP (if FI) | 18.1% | Table 7 |
| Supplement-interviewed HHs | 32,719 | Sample |
| HHs with food security status | 32,655 | Table 1 |

## Methodological Notes

- **Household assignment**: Race/ethnicity is measured at the household
  level from the reference person (PERRP = 40 or 41). This matches the
  convention used by USDA ERS in the annual food security report.

- **Hierarchy in `raceeth8`**: Hispanic ethnicity is assigned first (any
  race reporting Hispanic origin goes to `Hispanic`). Among non-Hispanic
  households, single-race reporters are assigned to their race; multiple-
  race reporters are classified as Multiracial NH. This matches the OMB
  SPD 15 convention for multiple-race tabulation.

- **AIAN alone, NH**: PTDTRACE = 3 (American Indian or Alaska Native
  alone). Does not include AIAN-in-combination; AIAN households that
  also reported another race are captured in `multi_group`.

- **Food security status**: Uses the pre-constructed HRFS12M1 variable
  (1 = High/Marginal, 2 = Low, 3 = Very Low). Households food insecure
  if HRFS12M1 >= 2; very-low food secure if HRFS12M1 = 3. Screened-out
  households are correctly classified as food secure.

- **Weights**: All estimates use household supplement weights
  (HHSUPWGT). Census stores weights as integers × 10,000 (4 implied
  decimals); the analysis divides by 10,000 to obtain actual weights.
  Weighted population counts use `sum(hhsupwtk)` (weight / 1,000), not
  `tab [aw=]`, which normalizes frequencies to the unweighted sample size.

- **Verification against ERR-358**: Aggregate weighted totals
  (134,062K households, 18,341K food-insecure, 13.7% FI rate) match
  ERR-358 Table 1A exactly. The four-category race/ethnicity FI rates
  (White NH 10.1%, Black NH 24.4%, Hispanic 20.2%, Other NH 11.9%)
  reproduce ERR-358 Table 2 to within rounding.

- **Reliability tiers.** Sample size varies by an order of magnitude
  across the groups reported. The table below groups every disaggregated
  cell into four reliability tiers so readers can judge how far to push
  each estimate. Design-effect-adjusted 95% CIs (deff = 1.5) are reported
  in Table 9 of the reproducibility `.do`.

  | Tier | N range | Approx SE (at p = 0.15, deff = 1.5) | Groups in this post |
  |------|---------|------------------------------------|----------------------|
  | A (publication-grade) | N ≥ 1,000 | ≤ 1.4 pp | White NH (N=23,101), Black NH (N=3,038), Hispanic total (N=3,951), Asian alone NH (N=1,681), Mexican (N=2,291) |
  | B (report-worthy) | 500–1,000 | 1.5–2.0 pp | Multiracial NH (N=504) |
  | C (interpretive) | 200–500 | 2.1–3.2 pp | **AIAN alone (N=272)**, Asian Indian (N=347), Chinese (N=376), Filipino (N=261), Other Asian (N=285), Puerto Rican (N=361), Central American excl. Salv. (N=271), South American (N=351), Other Spanish (N=262) |
  | D (directional only) | N < 200 | ≥ 3.3 pp | **NHPI alone (N=108)**, Japanese (N=150), Korean (N=133), Vietnamese (N=129), Cuban (N=158), Dominican (N=141), Salvadoran (N=116) |

  Tier C and D estimates are reported because the signal is too large
  to be explained by sampling noise (every Tier C/D point estimate lies
  several standard errors above or below the relevant comparison
  aggregate), but readers should not draw fine-grained between-group
  comparisons inside Tiers C and D. Pooling multiple CPS-FSS years
  (e.g., 2020–2024) would promote most Tier D cells to Tier B and is
  recommended for publication-grade tabulations. Note: because the
  CPS-FSS was cancelled in September 2025, the 2020–2024 window is now
  the **final available pooling window** for CPS-FSS race/ethnicity
  disaggregation.

- **MENA gap**: The December 2024 CPS still used the 1997 SPD 15
  two-question format with no MENA minimum category. Households with
  ancestry from North Africa or the Middle East are distributed across
  White NH, Other NH, and (rarely) multiracial depending on how they
  responded to the race question. Because USDA cancelled the CPS-FSS in
  September 2025 — with the December 2024 collection being the last —
  MENA food insecurity **will never be tabulable from CPS-FSS**. A
  MENA-specific rate will require a successor survey that adopts the
  2024 SPD 15 combined race/ethnicity question.

- **Asian subgroup sample allocation**: CPS allocates an Asian subgroup
  code (PRDASIAN) only when PTDTRACE indicates Asian alone or certain
  Asian-in-combination codes. Multiracial households that include an
  Asian component but were not classified as "Asian alone" are in
  `multi_group`, not `asian_detail`.

## Software

- Stata 14 or later
- No additional packages required

## References

Hales, L.J. & Coleman-Jensen, A. (2024). *Household food insecurity
across race and ethnicity in the United States, 2016–21* (Report No.
EIB-269). U.S. Department of Agriculture, Economic Research Service.
https://www.ers.usda.gov/publications/pub-details?pubid=108904
— Prior ERS disaggregation using 6 pooled years of CPS-FSS. Reports
AIAN, Asian, NHPI, multiracial, and Hispanic/Asian origin subgroup rates.
Directly comparable methodology (reference person, household weights).

Marks, R., Jacobs, P., & Coritz, A. (2023). *3.5 Million Reported
Middle Eastern and North African Descent in 2020*. U.S. Census Bureau
(America Counts: Stories Behind the Numbers, September 21, 2023).
https://www.census.gov/library/stories/2023/09/2020-census-dhc-a-mena-population.html
— Source of the 3.5 million MENA-descent figure cited in the MENA
data-gap discussion. Based on 2020 Census Detailed Demographic and
Housing Characteristics File A (DHC-A) write-in responses under the
White racial category (per the 1997 SPD 15 standard in force at the
time of the 2020 Census).

OMB Federal Register Notice (2024). "Revisions to OMB's Statistical
Policy Directive No. 15: Standards for Maintaining, Collecting, and
Presenting Federal Data on Race and Ethnicity." 89 FR 22182,
March 29, 2024.
https://www.govinfo.gov/app/details/FR-2024-03-29/2024-06469

Rabbitt, M.P., Reed-Jones, M., Hales, L.J., Suttles, S., & Burke, M.P.
(2025). *Household food security in the United States in 2024* (Report
No. ERR-358). U.S. Department of Agriculture, Economic Research Service.
https://www.ers.usda.gov/publications/pub-details?pubid=113622

U.S. Census Bureau (2024). *Current Population Survey, December 2024:
Food Security Supplement Technical Documentation*.
https://www.census.gov/data/datasets/time-series/demo/cps/cps-supp_cps-repwgt/cps-food-security.html

## Citation

Rabbitt, Matthew P. 2026. "Food Security Fridays — The Groups You Won't
Find in the Annual Report." Food Security Fridays series.
https://matthewpatrickrabbitt.com/food-security-fridays.html