US spending on health care as a whole has famously expanded as a share of the US economy over time, from 5% of US GDP in 1960 to about 20% of US GDP at present. However, total US spending on treatment services or mental health has remained at about 1% of GDP since 1975. Moreover, the share of Americans receiving mental health treatment has increased ove rtime. These patterns can be explained by the shifts in how mental health treatment is delivered. Richard G. Frank and Sherry A. Glied. 2023. “America’s Continuing Struggle with Mental Illnesses: Economic Considerations” (Journal of Economic Perspectives, 37:2, 153-78). (Full disclosure: I’m the Managing Editor of JEP, and have been so for 37 years now.)
There are four main reasons for the difference in spending growth between mental health care and general medical care.
First, the main driver of cost growth in the general health care sector has been technological change, particularly through the introduction of capital-intensive devices and procedures (Chernew and Newhouse 2011). In contrast, the technology of treatment in mental health continues to rely on labor and prescription drugs. Newer treatments for mental health conditions have typically offered few gains in efficacy, although they have generated improvements in treatment adherence and outcomes by reducing side effects and increasing the tolerability of treatments (Insel 2022). While psychopharmacology experienced considerable innovation prior to 2000, relatively few new classes of drugs for treating mental
illnesses have been introduced since then. …
Second, over the past 50 years there has been dramatic, cost-reducing substitution for the human and institutional inputs that were previously used to provide mental health care. In 1975, 63 percent of mental health care spending was for institutional care in hospitals and nursing homes; today, 31 percent of expenditures occur in these costly settings (SAMHSA 2014; 2016). Treatment with prescription drugs has taken a central position in treatment of mental illnesses, often substituting for costlier psychotherapy for the most prevalent mental health conditions, depression and anxiety. … The cost of psychotherapy itself has also dropped sharply because the mental health sector has been far more accommodating of diverse types of health care providers than has general health care. Psychotherapy provision has shifted from treatment by psychiatrists and PhD-level psychologists to treatment by social workers, counselors, and MA-level psychologists. … Today over 90 percent of psychotherapists are trained below the doctoral level, a far higher share than in the 1970s and 1980s. The shift towards lower cost professionals with less extensive training has driven the costs of psychotherapy down, without any documented evidence of a reduction in quality—although no recent studies have directly compared the quality of services delivered by those with varied professional training …
Third, a much larger share of mental health care (just under two-thirds) is paid for by public funds (about one-third is paid by Medicaid) than is the case for general health care, and a much larger share—20 percent—is paid for by programs under fixed budgets. Public programs generally pay lower prices. …
Finally, mental health spending appears to be growing much more slowly than general health spending, in part because of a change in classification. In the 1970s and 1980s, when institutional treatment of those with serious mental illness accounted for a much larger share of mental health spending than it does today, all the expenses of institutional treatment—including the costs of whatever limited clinical treatment was provided as well as the costs of institutional room and board, often of poor quality—were counted as part of mental health spending. Today, the costs of housing and food for people with serious mental illness, who are not typically institutionalized, are no longer counted as part of mental health treatment
The authors also point out that support and services for mentally ill people end up being provided in a range of non-health-care contexts. They draw up on a wide array of evidence across studies and programs to compile the following table.
As the authors point out, it seems plausible that the US is investing too little in care for those who have serious mental illnesses, but too much for those with milder concerns. They write:
Current policy choices have led to a misallocation of resources in the delivery of clinical services. Too few people with treatable mental health conditions, including those with serious illness, obtain care that could help them. This situation may arise, in part, because the decisions of people suffering from mental illness to seek care may not accurately reflect the likely value of such care to themselves and to others, as well as because of underinvestment in treatment capacity for the most serious conditions. At the same time, moral hazard associated with insurance coverage of mental health services may lead to overuse (or inappropriate use) of some services within this category, either to address problems of living that cause relatively little impairment or because the quality and nature of treatments are so variable. Both overuse and underuse reflect the fundamental difficulty of matching people and treatments in the face of great heterogeneity and uncertain diagnosis.