Mental Health and Addiction Care Is 7% of US Health Spending

Treatment for mental health conditions and addiction accounted for $174 billion a year in the United States between 2021 and 2023 — about 7% of all medical spending — according to an analysis of federal survey data published Sept. 30 in JAMA Psychiatry. The study's authors, led by Carlos Blanco of the National Institute on Drug Abuse, concluded that the figure is strikingly small relative to the toll those conditions take on the population.

The research drew on the Medical Expenditure Panel Survey, an annual, nationally representative sample of the civilian, noninstitutionalized U.S. population that the Agency for Healthcare Research and Quality uses to track who pays for care and how much. The analysis covered 67,532 person-year observations, of which 15,407 involved people who reported at least one outpatient mental health visit.

What the survey captured

Total health care expenditures in the three-year window came to an estimated $2.5 trillion annually, with a confidence interval running from $2.3 trillion to $2.7 trillion. Mental health expenditures were estimated at $174 billion, with a confidence interval of $152 billion to $196 billion — a range that puts the share of overall medical spending at 7.0%, and at worst 7.7% even at the high end of the estimate.

Blanco's co-authors were Chandler McClellan and Samuel H. Zuvekas of the Agency for Healthcare Research and Quality and Mark Olfson of Columbia University. The authors reported that data analysis was conducted between August 2025 and January 2026.

Because the survey asks people what care they actually received and paid for, the totals measure the cost of treated illness. They do not include the cost of conditions that never reached a clinician, which is where a large share of substance use disorder sits: the Substance Abuse and Mental Health Services Administration's National Survey on Drug Use and Health has found year after year that only a minority of people who meet the criteria for a substance use disorder receive any treatment.

Where the dollars concentrate

Two diagnostic categories dominate the aggregate. Mood disorders accounted for $74 billion and anxiety and fear-related disorders for $69 billion, with confidence intervals of $62 billion to $86 billion and $58 billion to $81 billion respectively. Together they represent more than four-fifths of the total.

The spending also skews heavily toward outpatient care. Ambulatory services accounted for $107 billion, while prescription medication came to $36 billion, home health care $20 billion and inpatient treatment $12 billion — a distribution that reflects how mental health care is delivered in the United States, mostly in offices, clinics and pharmacies rather than hospitals. It also reflects the structure of the levels of care that insurers are most willing to pay for.

On a per-person basis, the numbers look modest at the population level and considerably larger among people in treatment. Average annual mental health spending across the whole sample was $523 per person. Among those with any treated mental health disorder, it was $2,473.

Addiction costs the most per patient

The most striking figure in the study sits in a table rather than the headline: measured per treated individual, substance use disorder was the most expensive condition in the analysis, at $9,569 per person. Schizophrenia followed at $9,370 and bipolar disorder at $7,382.

That ranking cuts against the common intuition that addiction is a cheap condition to treat because the medications — buprenorphine, methadone, naltrexone — are generic and inexpensive. The gap is explained partly by the settings involved. Addiction care often runs through emergency departments, detoxification units, inpatient stays and repeated short-term episodes rather than a single continuous course of treatment, and the analysis counts the full range of medical services a person uses, not just the addiction-specific ones.

The comparison also carries a policy implication the authors do not belabor: if a relatively small number of patients account for a disproportionate share of spending, then the design of the care they receive matters a great deal.

Why the authors call it low

Blanco and his colleagues close by arguing that the expenditure data look low "relative to the societal burden of mental disorders." That framing matches a wider body of work on the economics of mental illness, which consistently finds costs far larger than what the health system itself spends once lost productivity and premature death are counted.

The authors point to three levers they say could change the picture without simply increasing volume: research into more effective treatments, reductions in low-value care that does not help patients, and better matching of a patient's impairment to the appropriate level of care. In practice, that last item means steering people toward the least intensive setting that can hold them rather than defaulting to whichever bed happens to be open.

The study is not a clinical trial and offers no new therapy. Its value is descriptive, and its main contribution is a decomposition that shows where mental health and addiction dollars sit inside American medical spending — a baseline from which any future change can be measured.