who funds the organizations that run mental health care: the psychiatric and psychological apas, the ama, samhsa and nimh

five organizations sit between a person and a psychiatric diagnosis, a prescription and a bill. one writes the manual. one owns the billing codes. one licenses the research literature. one funds the services. one funds the science. this page is what each of them lives on, from its own filings.

twelve sources: the american psychiatric association's irs return, the american psychological association's audited statements, the ama's audited annual report, the federal regulation that makes the ama's codes compulsory, the ama's own licensing page, three peer-reviewed studies of the government's open payments database, the senate appropriations summary of the 2026 bill, two congressional research service reports and nimh's own budget request. every percentage is our arithmetic on their lines. the fda, the sixth institution, has its own page.

the american psychiatric association: a publisher with a meeting

the body that writes the dsm is a 501(c)(6) business league with 248 employees, 20 trustees and, in 2023, $64.6 million in revenue^[1]. its return breaks the money out like this:

put the publishing lines together and a third of the association is the sale of books and journals, which is to say the dsm, its text revision and the journals that carry the literature. a fifth is the annual meeting. members' dues are eight percent^[1]. the five largest contractors it paid that year were an audio-visual firm, two annual meeting services companies, an e-publishing vendor and a printer, which is the same story from the expense side^[1]. officers and key employees were paid $5.8 million in total; 78 people earned over $100,000^[1]. in 2024 revenue fell to $58.2 million against $60.9 million of expenses, and the incoming chief executive was paid $462,395 for a part year, her predecessor $749,990 for the full year before^[1].

the leaders. a 2020 bmj study matched 328 leaders of the ten us medical associations covering the costliest disease areas against open payments, the federal database of what manufacturers pay physicians. 72 percent had a financial tie; the total over 2017 to 2019 was about $130 million and the median per leader $31,805. the american psychiatric association's leaders had a median of $212, against $518,000 at the society of clinical oncology and $251,464 at the college of rheumatology^[2]. the authors' summary: “Financial relationships between the leaders of influential US professional medical associations and industry are extensive, although with variation among the associations.”^[2] psychiatry's leadership sat at the low end of that variation.

the members. the profession as a whole did not. of 56,955 us psychiatrists, 75.0 percent received at least one non-research industry payment between 2015 and 2021, $357,971,774 in total, and “1% of psychiatrists received 74.7% of industry payments”^[3]. research money is a separate and smaller stream: 488 psychiatrists, 0.85 percent, were principal investigators on industry research support totalling $98,189,901 over 2015 to 2022, median $21,749 a year, and “payments were highly concentrated among a few companies and psychiatrists from a few institutions”^[4]. the same concentration is what the funding entries on this site's doctors' education page show from the other side.

the american psychological association: a database licensor with a landlord's income

the psychologists' association is twice the size of the psychiatrists' and is a different kind of business. its 2024 audited statements report $143.0 million in revenue^[5]:

two-thirds of the association is licensing: the notes describe “annual licenses of electronic products” as almost three-quarters of what it is owed at year end, and the products are the psycinfo family of databases that every university library and many hospitals pay for^[5]. the statements' own summary is plainer than the table: “APA's operations are supported primarily by membership dues, subscriptions, and publication revenue.”^[5] dues are six percent of that sentence. the association also owns two washington office buildings through subsidiaries, one of about 370,000 square feet and one of about 263,700, which brought in $18.9 million of rent against $20.7 million of building costs and $4.2 million of interest on $105 million of long-term debt^[5]. it spent $143.96 million, more than it took in from operations, and finished the year with $54.5 million in net assets after investment returns^[5]. its political arm, apa services, takes 60 percent of every dues dollar and runs a federal political action committee^[5].

the ama: a royalty on a mandate

the american medical association is the largest of the three and the one whose money is least about medicine. its 2024 audited statements report $513.2 million in revenue^[6]:

the royalties are mostly cpt, current procedural terminology, the five-digit codes on every medical bill in the country. the ama's own licensing page: “CPT is copyrighted by the AMA and as such must be licensed to be used in electronic products or other forms”, and “Any entity using CPT content anywhere worldwide must have a license authorizing the use of CPT code set from the AMA or an authorized distributor.” royalties are set “based on the type of product in which the CPT content is used and the type of user of the product”^[8].

what makes the licence a mandate is not the ama but the federal government. the hipaa code-set regulation adopts “Current Procedural Terminology, Fourth Edition (CPT-4), as maintained and distributed by the American Medical Association” as the standard for “physician services” and a list of others, from physical therapy to ambulance transport, that every covered claim must use^[7]. a psychiatrist billing a 45-minute psychotherapy session, a psychologist billing a testing hour, an insurer processing either: each is using a code set the regulation names and the ama owns. the royalty line has grown from $308.0 million to $326.0 million in a year; dues fell^[6].

the rest of the picture: $279.4 million in compensation and benefits, $55.2 million of operating results before tax, a further $79.7 million of investment gains, and $1,150.7 million in association equity at year end, up from $1,023.4 million^[6]. the auditor is deloitte^[6].

hhs: the services agency that survived, and the research institute that stood still

the three associations above run on what they sell. the two federal bodies below run on appropriations, and in 2026 both were proposed for abolition as separate entities and both survived.

samhsa. the senate appropriations committee's summary of the enacted 2026 bill: “The bill provides $7.4 billion for the Substance Abuse and Mental Health Services Administration (SAMHSA), an increase of $65 million over fiscal year 2025”, and “In addition to rejecting President Trump's proposed $1 billion cut (-15%) to SAMHSA programs, the bill maintains SAMHSA as its own, independent agency”^[9]. inside it: “$991 million, a $5 million increase over fiscal year 2025, for the Mental Health Block Grant”; “$2.0 billion, a $5 million increase, for the Substance Use Prevention, Treatment, and Recovery Services Block Grant”; “$1.6 billion for State Opioid Response grants”; and “$535 million, a $15 million increase over fiscal year 2025, for the 988 Suicide Prevention Lifeline”^[9]. the congressional research service records the plan that was rejected: the budget “would have consolidated into a new Administration for a Healthy America (AHA) two existing operating divisions”, hrsa and samhsa, and “These reorganization proposals were largely not adopted in the House committee bill, Senate committee bill, or in the enacted FY2026 LHHS omnibus.”^[10]

nimh. the institute describes itself as “the lead federal agency for research on mental illnesses, with a mission to transform the understanding and treatment of mental illnesses through basic and clinical research”^[12]. its money has not moved. the crs table gives $2,191.7 million in fy2024, $2,191.7 million in fy2025 and $2,189.8 million in fy2026, a change of minus 0.1 percent, inside an nih total of $47.49 billion^[11]. the fy2025 request had asked for $2,548.7 million and did not get it^[12]. the fy2026 request went the other way and proposed a “National Institute of Behavioral Health” formed by consolidating three institutes, part of a plan in which “Four ICs would have been eliminated, while 19 would have been consolidated into eight restructured ICs”; “The FY2026 appropriations maintained the preexisting structure of NIH and did not adopt the proposed restructuring from the President's budget.”^[11] in fy2023, the last year with a full justification, nimh funded 2,430 research project grants worth $1,599.1 million^[12].

what it adds up to

the three membership bodies are not membership-funded. dues are eight, six and six percent^[1],[5],[6]. each lives on a product: the manual and the meeting, the database, the codes. that is not a scandal; it is what a professional association in the united states is. it does mean that the organization that decides what counts as a disorder earns a third of its income from selling the book that says so^[1], the organization that decides what a service is called earns two-thirds from licensing the names^[6],[7], and the organization that publishes the evidence earns two-thirds from selling access to it^[5].

the drug industry's money runs through the people more than the institutions. the psychiatric association's leaders had almost none of it by the government's own records^[2]; three in four practising psychiatrists had some, and a hundredth of them had three-quarters of it^[3]. the federal side is the reverse case: taxpayer money, flat for the science and held level for the services, in a year when the executive proposed shrinking both and congress declined^{[9],[10],[11]}.

what to ask

these are questions, not advice.

• what is the cpt code on my bill, what does it describe, and does it match what happened in the room?^[7]
• is my clinician listed in open payments, and for what? the database is public and searchable by name.^[3]
• which edition of the dsm was my diagnosis made under, and what changed in the criteria between editions?^[1]