Medical Costs Rose Most Rapidly for the Very Old and the Very Young

"In 1963, per capita spending on infants was only 53 percent of per capita spending on 35-44 year olds. By 1987, per capita spending on infants was 2.3 times as high as spending on this group. ..in 1963, per capita medical spending on those aged 85 or more equaled per capita medical spending for 35-44 year olds... by 1987, spending per person aged 85 or more was 5.2 times spending per 35-44 year old."


In the United States, as in most high-income countries, annual medical spending has grown a few percentage points faster than income for over three decades. In 1987 dollars, medical spending in the United States rose from $278 per person in 1953 to $1,521 per person in 1987. Real spending grew fastest in 1963 to 1970, averaging 7.9 percent per year. It was during that period that Medicare and Medicaid were introduced and private insurance coverage expanded as well. On average, between 1953 and 1987, medical spending grew annually by 5 percent per year.

For which groups of the population is medical spending growing fastest? And, has this increased spending been accompanied by rapid improvements in health? In The Medical Costs of the Young and Old: A Forty Year Perspective, (NBER Working Paper No. 6114), NBER Research Associate David Cutler and Ellen Meara show that medical spending grew fastest for the young (those less than one year old) and the old (those aged 65 or more). Between 1963 and 1987, real spending on infants grew by an average of 9.8 percent a year; spending on those aged one to 64 grew by 4.7 percent a year; and spending on the old rose by 8 percent a year. In 1963, per capita spending on infants was only 53 percent of per capita spending on 35-44 year olds. By 1987, per capita spending on infants was 2.3 times as high as spending on this group. Also noteworthy is the high growth of spending on the "oldest old." Whereas in 1963, per capita medical spending on those aged 85 or more equaled per capita medical spending for 35-44 year olds, by 1987, spending per person aged 85 or more was 5.2 times spending per 35-44 year old.

Almost all of the difference in growth rates of spending for young and old was accounted for by high growth in spending for a small percentage of each group. For example, just 10 percent of infants accounted for almost all of the difference in spending growth between infants and those aged one to 64. Similarly, high spending growth for the most expensive 10 percent of the elderly accounted for 66 percent of the excess spending growth for the elderly. For the top 10 percent of the elderly, the most common diagnoses were circulatory system disorders, accounting for 28 percent of this group.

The good news is that gains in health outcomes for both old and young are consistent with a fair return for this higher spending. The age-adjusted death rate attributable to heart disease, for example, fell from 287 per 100,000 in 1960 to 152 per 100,000 by 1990. However, the consensus among medical researchers is that technologically intensive medicine has been less important to improved health than life-style changes and pharmaceuticals. For infants, on the other hand, it is more likely that health gains reflect additional spending on medical technology, such as neonatal intensive care for low birth weight babies. Between 1960 and 1991, for every thousand babies weighing 2.2 pounds to 3.3 pounds at birth, deaths dropped 80 percent, from 549 to 91. In other words, the odds that such a low-weight baby would survive improved from slightly worse than even to better than 9 in 10. Similarly, death rates for babies weighing 3.3 pounds to 4.4 pounds fell 75 percent, from 207 per thousand to 40.

Cutler and Meara point out that most forecasts of medical spending fail to account for changes in relative spending by age. They conclude: "Our results suggest this understates the future growth of costs, since the fastest growing age group, the 85+ population, is also the group whose costs are growing most rapidly."


The Digest is not copyrighted and may be reproduced freely with appropriate attribution of source.