The United States reached a record real median household income of $87,460 in 2025, according to data released by the Census Bureau on September 15, 2026. The official poverty rate also fell to 10.2 percent. These are important gains. They do not show that economic insecurity has disappeared or that public benefits have become unnecessary.
The same release reports that 34.5 million people remained below the official poverty line. The Supplemental Poverty Measure was 13.1 percent, a rate that was not statistically different from 2024. Medicaid coverage declined, and 26.7 million people were uninsured for the entire year. Social work needs to communicate the entire picture rather than choosing the most reassuring number.
Two measures answer different questions
The official poverty measure compares pretax cash income with a national threshold that varies by family size and composition. In 2025, the weighted average threshold for a family of four was $32,970. The measure offers a long historical series, but it excludes many benefits and does not adjust the threshold for geographic differences in housing costs.
The Supplemental Poverty Measure uses a broader accounting. It includes taxes and tax credits, many government benefits, geographic differences in housing expenses, work costs, and medical expenses. It does not replace the official measure. It reveals how public policy and unavoidable expenses change the resources a household can actually use.
That distinction matters for practice. A family may be above the official threshold and still be unable to pay rent, transportation, child care, medication, or utility bills. A national income gain may coexist with severe local hardship.
The safety net is visible in the results
The Census Bureau estimates that Social Security moved 28.8 million people out of poverty under the Supplemental Poverty Measure in 2025. That is not evidence that recipients are dependent on an unnecessary program. It is evidence that a program accomplished its central purpose.
Social workers should apply the same logic to other benefits. A lower poverty rate after assistance does not prove that the assistance can be removed without consequence. It shows that the assistance is part of the household’s resources. Evaluating a safety net requires asking what happens with the benefit and what would happen without it.
This is especially important when public discourse separates those who receive aid from those who contribute. Most households participate in public systems across the life course. They use public schools, tax benefits, health coverage, retirement insurance, disability protection, housing infrastructure, and emergency services at different times. A serious analysis asks whether those systems reduce hardship efficiently and fairly.
Health coverage moved in the wrong direction
The share of the population with Medicaid coverage decreased by 0.5 percentage points in 2025. For adults ages 19 through 64, Medicaid coverage fell by 0.6 percentage points. The Census Bureau also found that 7.9 percent of people, or 26.7 million, were uninsured for the full year.
These figures should prompt service agencies to examine enrollment loss, renewal problems, and the affordability of care. A person who loses coverage may delay treatment, ration medication, accumulate debt, or rely on emergency care. Each result can increase the workload of housing, disability, child welfare, behavioral health, and crisis programs.
National improvement can hide unequal conditions
Median income represents the income of the middle household. It does not show the experience of every group or region. The Census Bureau reports that household income at the 90th percentile increased while income at the 10th percentile did not change significantly. State comparisons also show that the Supplemental Poverty Measure exceeded the official measure in 26 states and the District of Columbia when using three-year averages.
Social work organizations should pair national figures with local evidence. Useful measures include eviction filings, food pantry demand, utility arrears, benefit processing time, health coverage loss, transportation barriers, and the number of people who abandon an application before completion.
Agencies should examine what their own data miss
A waiting list counts people who reached the agency. It does not count everyone who could not find the program, lacked documents, could not complete a digital form, or stopped trying after repeated denials. Agencies should gather information about these points of exclusion and publish corrective steps.
They should also distinguish outputs from outcomes. A referral, application, or benefits screening is an activity. The outcome is whether the person obtained food, housing, health care, income, or another needed resource and whether the improvement lasted.
A reform agenda grounded in the new data
Social work leaders should respond to the 2025 findings with several commitments:
- Use both poverty measures when explaining economic need.
- Describe the antipoverty effect of benefits before proposing cuts.
- Track Medicaid loss and help eligible people resolve enrollment barriers.
- Publish local service access and completion data, not only referral totals.
- Include housing, medical, transportation, work, and child care costs in assessments of hardship.
- Challenge language that treats benefit receipt as a moral failure rather than a policy condition.
A record median income is good news. The professional responsibility is to interpret it accurately. Millions of people remain poor, public programs are already preventing deeper poverty, and health coverage has declined. Reform should build on what works while correcting gaps, not use one favorable statistic to erase the people whose lives the median cannot describe.
Leave a Reply