The RAND Health Insurance Experiment is the best long-term study of the effects of insurance on spending and quality of care. Between 1974 to 1982, RAND studied 7,700 in six cities. The experimental group received free health-care and the control group shared costs with an HMO. They compared the spending on treatments.

So how much more did the free-care group spend and how much did their health improve?
The main finding was a lack of correlation between increased spending and increased health.

More isn’t better. Spending more beyond a point doesn’t increase lifespan or quality of life. In some cases, too much makes things worse. Look how overuse of antibiotics creates bacteria resistance.

The RAND study found that spending increased between 30%-40% in the free-healthcare group, but more spending did not improve health. Those in the free-care group visited the doctor and hospitals more often and received more treatments than the cost sharing group. The best improvements were for eyeglasses (which isn’t even medicine). A peer-review by doctors found that 23% of the treatments given were inappropriate for both the test and control groups.

Rand has a bibliography of over 300 papers relating to the original research. Some made similar but more limited studies. Aggregate correlation studies find the same result.

Likewise, cross-country studies show no relation between increases in health-care spending and health quality or mortality rates. Europeans spend roughly 10% or less of their GDPs on healthcare compared the the Americans 16%. There’s very little difference in health quality. This lack of correlation holds for Asia, Africa, and South America.

Most medicine is not very effective at the margins and it does not appear that extensive care improves health much at all.

The intitial spending has the greatest benefit. Afterwards, the marginal utility of each unit of health care produces less and less effect. The greatest benefits in medicine reduced infant mortality and eliminated epidemics with vaccines. Spending beyond a certain point accomplishes nothing: this is especially obvious when treating end of life patients during the 6 months. The costs increase exponentially without increasing lifespan.

Antibiotics are example of marginal effects that turn negative. The first pill has the greatest marginal effect, the second pill has a good but more moderate effect, and so on. The 30th pill has very minor positive effects and raises the negative effects – it kills too much of good symbiotic bacteria and causes the bad bacteria to develop a resistance. Taking 500 pills is negative.