From the 1970s: A Rand study on health insurance economics
In the early 1970’s, the federal government commissioned the Rand Corporation to study the effects of cost sharing . This was truly a randomized, controlled, economic experiment in which over 5,000 families were randomly assigned to receive health insurance plans with levels of 0, 25, 50, or 95 percent cost sharing. All plans had the same maximum out of pocket spending limits which were adjusted according to pre-set standards based on income for poor participants. Rand followed the families for three to five years.
The study revealed that cost sharing is likely to result in a mix of desirable and undesirable outcomes. Among the most unpleasant for proposnents of HSA/HDHP’s is that cost sharing did not improve the efficiency of the health care market. The cost per episode of health care utilization was no different among the groups, nor was the incidence of inappropriate hospitalization or antibiotic usage. Instead of increasing efficiency, cost sharing served to decrease the frequency and quantity of health care utilization. Screenings were more common among the group receiving free care, and both mental health and well visits were negatively correlated with cost sharing. Rand concluded that cost sharing led to worse control of hypertension, worse vision care, and worse oral health.
On the other hand, Rand found that health care spending was undeniably reduced through cost sharing plans. Cost sharing had less effect on acute care than on other areas of health care, and the value of the increased care that was sought with free plans may have been only marginal. Free care recipients had more self-reported diseases but overall, participants in the study reported no significant difference in self-assessed measures of health. The differences reported for hypertension management were actually the result of detection, not care. Once the disease was diagnosed, management and follow-up were similar among all groups. Cost sharing resulted in no overall difference in provider choices or the quality of care received.
The insurance plans in the study included no such savings accounts and no associated tax incentives, so we cannot know exactly how HSAs may have affected the results. The differences in spending habits and health outcomes between the groups with various levels of cost sharing are smaller than those between insured and uninsured groups; consequently, increasing access to insurance through increased cost sharing is likely to be beneficial to society as a whole. Thus, while certainly demonstrating areas of weakness and the need for improvement from HDHP’s alone, the Rand Insurance Trial may give some backing for the ideals motivating HSA’s as well as the combination of HSA/HDHP’s.
