Each September, the Census Bureau publishes the country's two answers to the question of how many Americans are poor, and the two answers disagree by design. The official poverty measure, computed since the 1960s, compares a family's pre-tax cash income to a threshold scaled for family size. The Supplemental Poverty Measure, produced since 2011, counts taxes, in-kind benefits, and necessary expenses, and adjusts the threshold for housing costs. In recent years the two measures have diverged sharply, with the SPM documenting both the power of refundable tax credits and food assistance and, in the years after pandemic programs expired, the fastest single-year increases in child poverty on record.
Why do the two measures exist?
The official measure was built around a 1955 observation that families spent about a third of their income on food, multiplied by three, and updated for inflation ever since. It has never accounted for the programs that constitute the modern safety net: SNAP, housing vouchers, the Earned Income Tax Credit, and the Child Tax Credit arrive as benefits or credits, not cash paychecks, so the official measure literally cannot see them. The SPM was designed by an interagency working group to fix that, and its annual publication, alongside the official rate in the Income in the United States and Poverty in the United States reports, has become the authoritative record of program effects.
What did the pandemic years document?
The 2021 SPM delivered the clearest natural experiment in modern anti-poverty policy. With expanded Child Tax Credit checks, stimulus payments, and enhanced unemployment insurance, the SPM child poverty rate fell to its lowest recorded level, 5.2 percent, while the official measure barely moved. When the credit expired at the end of 2021, the 2022 SPM child rate more than doubled, from 5.2 to 12.4 percent, the largest one-year increase in the series, a rebound documented in Census tables and analyzed across the economics literature. No labor-market change of comparable speed occurred in either direction; the swing was, on the record, the programs.
What did the most recent reports show?
The September 2025 Census reports, covering income year 2024, documented a labor market doing the heavy lifting: real median household income rose meaningfully, the first significant increase in several years, driven by gains at the bottom of the wage distribution as low-wage pay growth outpaced inflation. The official poverty rate edged down from its post-pandemic plateau, while the SPM, more sensitive to program retrenchment and housing costs, told a more complicated story, with poverty among children and among renters remaining above pre-pandemic SPM baselines even as the labor market strengthened. The divergence is the point: a country can have falling official poverty and rising rent burden simultaneously, and both numbers are correct on their own terms.
Why the measurement fight matters now
Statistical infrastructure has itself become news. The 2025 federal funding environment cut and shortened several Census surveys, and USDA's September 2025 termination of the food-security supplement, reported in department announcements, removed a companion series that had run since 1995. Advocates and researchers documented the risk: poverty measurement shapes program formulas, refundable-credit parameters, and school-meal eligibility thresholds. A measurement cut is a policy change wearing a lab coat.
How should a reader use these numbers?
Three rules of thumb hold up. Use the official measure for long historical comparison, since its definition is frozen. Use the SPM to ask whether programs are working, since it is the only series that counts them. And never accept a single-year move in either series as a trend, because the survey's sample and thresholds shift, and both series are revised. The honest picture from the current record is a low-wage labor market that raised pay faster than prices, an expiration of pandemic-era supports whose absence shows in the supplemental series, and a measurement apparatus under budget pressure. Two numbers, two Americas, and the distance between them is the safety net, drawn to scale.
For more context, read The Last Food-Insecurity Report: 13.7 Percent, Then Silence.
For more context, read economic mobility by county.
For more context, read The Generation That Bought Late or Never.