True Price of Depression Treatment is Declining

More could be done to treat major depression in 1995 than in 1991 for a given budget.

After health expenditures rose at an alarming pace for much of the post World War ll period, the increases in the 1990s have been relatively anemic, much to the relief of policymakers, business managers, and health professionals. But what accounts for this welcome trend?

Economists typically start delving into the forces driving trend change by dividing nominal spending by an official price index. Longstanding difficulties in the construction of price indexes in the health sector have been an impediment to gaining a deeper understanding of health care productivity and spending. For example, the CPI only deals with consumers' direct out-of-pocket payments; it doesn't include any money paid by employers to insurers. The BLS's medical price indexes also are based on the repricing over time of a fixed basket of goods or services, rather than capturing the entire treatment of an illness and condition and the impact of technological and institutional innovations.

In a recent NBER Working Paper, Price Indexes for Acute Phase Treatment of Depression (NBER Working Paper No. 6799) , authors Ernst Berndt, Susan Busch, and Richard Frank have created a price measure for the treatment of one major illness: depression. One in ten American suffers from major depression. The authors identify therapeutically similar treatment bundles, based on an intensive review of the clinical literature. They also tap into a medical claims database of almost half a million lives, annually from 1991 to 1995. The price indexes they calculate are based on transactions prices rather than list prices; they take into account treatment episodes, quality, and expected outcomes; and, finally, they use time-varying expenditure weights.

The researchers also distinguish between a supply-side index and a demand-side index. Their supply side measures, like the BLS's producer price index, represent the total receipts received by providers of medical treatment from both the insurer and patient. Their demand side index, like the CPI, incorporates the out-of-pocket payments by consumers or patients.

The results are striking. All the supply-side indexes are flat or decreasing over the 1991 to 1995 period. In sharp contrast, the BLS producer price indexes point to sharply rising prices. The authors' demand-side index shows annual increases of about 10 percent between 1991 and 1994, and then a 20 percent surge in 1995. The reason: the impact of higher co-payments and deductibles.

Taken altogether, the real price of care has fallen: more could be done to treat major depression in 1995 than in 1991 for a given budget. But contrary to both public and expert perception, increases in spending on mental health care have not been driven by more low benefit services and higher payments to providers. "Our results point to a different story where spending increases are due to a larger number of 'effective' treatments being provided," the authors write.

-- Chris Farrell