Sam Robinson
Healthcare costs grew by over 6,500% between 1970 and 2023. Decades of research have provided valuable insight into excessive healthcare spending, yet few studies have made contributions as sizeable as the RAND Health Insurance Experiment (HIE).
Conducted from 1974 to 1981, this groundbreaking government-sponsored study was designed to evaluate the relationship between health insurance, costs, and demand. Not only is it the largest US healthcare study conducted to date, it was also a randomized control trial (RCT)—yielding robust evidence while establishing the feasibility of controlled experiments in healthcare economics.
The study enrolled 7,791 U.S. participants across 6 sites, with no individuals being eligible for Medicare or service-related health benefits. Each participant was enrolled in one of six care plans:
Plan participants were evaluated for health status across several health markers, and healthcare spending was recorded based on visit frequency and expenditure. Relationships between plan type, utilization, and health status were then evaluated to draw conclusions around the following questions:
The following are some of the key takeaways.
Cost-sharing reduces likelihood of use, not the amount spent per medical encounter. Compared to the 95% coinsurance plan, the free plan had 67% higher outpatient expenses and a (nearly identical) 66% higher outpatient visit rate. Visits and costs changed proportionally, signifying an increase in utilization without a decrease in per visit costs. Inpatient care followed a similar pattern, with the free plan having 30% higher expenses on 29% more visits compared to the 95% coinsurance plan.
The probability of overall medical service use increases with income.
Higher income is linked to greater overall utilization and increased outpatient service consumption but is associated with decreased inpatient care utilization.
This can be attributed to a lower cost-sharing limit for the poor. Those with less income had lower out-of-pocket maximums and qualified for free inpatient care sooner, while those who were assigned a $1,000 maximum tended to use cheaper outpatient services and remained below the limit.
The HMO group had a 35% reduction in inpatient admissions and 28% lower expenditure than the free FFS plan.
By comparison, the family pay plans achieved a more modest 15-20% reduction.
When the results of the experimental HMO group’s reduction were compared to an established HMO, there was no significant difference between the two—indicating that the observed change in expenditure reflects a real-world phenomenon.
It is important to note that individuals placed in the experimental HMO group were less satisfied overall compared to the fee-for-service and pre-existing HMO groups, which had similar levels of satisfaction.
Finally, an evaluation of health markers revealed that free care for the poorest group of patients saw reduced hypertension, improvements in vision and oral health, and a reduction of general serious symptoms. By contrast, free care did not result in any clinically significant benefits for the average individual.
How are these results best interpreted?
Although the US healthcare system has undergone substantial change since this study was published, human decision-making in medical spending remains largely the same, so the main takeaways from this project are still applicable today.
Prior to this study’s publication, it was commonly argued that medical care was price inelastic, meaning that an individual’s consumption of medical care was driven predominantly by need rather than choice. The HIE was instrumental in rejecting this idea by establishing that insurance-induced changes in price influence the demand for healthcare goods. This concept, known as the “moral hazard” of health insurance, suggests that individuals are more likely to consume (and overutilize) medical services when they are protected from the financial consequences of doing so.
The insurance moral hazard is often invoked as a criticism of more affordable or subsidized healthcare, based on the idea that decreased out-of-pocket cost leads to excessive use that drives higher prices. While there is truth behind this idea, the reality is not as straightforward. Human consumption habits are much more complex, and the relationship between cost-sharing and consumption is not as direct.
In spite of this nuance, there may still be opportunities for innovation. Variations in the type and quantity of out-of-pocket spending could be used to effectively allocate healthcare resources. Consumption could be discouraged in healthier populations by using increased cost-sharing and encouraged for higher-risk groups by minimizing out-of-pocket expenses for covered services. Nonetheless, the efficacy of such interventions requires further analysis of the relationship between costs and consumption.
The HIE did not demonstrate an effective application of cost-sharing. Simply giving patients “skin in the game” did not make them smarter buyers—participants with higher cost-sharing levels did not achieve greater value for their healthcare dollar but instead spent less for proportionally fewer healthcare services. While cost-sharing did not function effectively as a standalone tool, this doesn’t reject the possibility that it can be useful alongside additional components like price transparency and open-access efficacy information. The technological age has created new opportunities for growth in both of these areas.
The decreased satisfaction seen in the experimental HMO enrollees reflects one of the biggest flaws of HMOs—consumers don’t like being forced to accept limited options. This lack of choice was a driving factor behind the rejection of HMO’s during the late 1990s in favor of more flexible preferred provider organizations (PPOs). Healthcare leaders should recognize the relationship between consumer choice and satisfaction when attempting to design sustainable systems.
Policymakers should appreciate the clinical benefits that free care provided to the poorest echelon of enrollees. Targeted free care contributed to significant improvements across several key health markers. Similar interventions for marginalized populations can reduce spending and improve resource allocation by addressing illnesses before they escalate. For example, treating hypertension, which affects nearly half of U.S. adults and kills over half a million Americans every year, is one of the most cost-effective methods for reducing cardiovascular complications—a signal that broad health interventions may serve as a means for reducing cost.
For academics, the RAND HIE demonstrates that, with proper planning and resources, controlled experimentation in healthcare policy is possible. As our healthcare system becomes more convoluted, researchers and policymakers must work together to make decisions that will define healthcare for our future generations. The HIE offers a paradigm of collaboration between government and researchers, whose combined resources and intellect can drive the development of informed solutions to complex problems.
CMS HCC-28 is vital for Medicare Advantage risk adjustment, aligning payments with enrollee health needs. How does this impactful model refine patient data and shape today’s healthcare landscape? Learn more about the model’s updated framework and its impact on accurate reimbursement.
How has the 340B Drug Discount Program influenced healthcare? In what ways has a program built for the poor come to serve the rich, and what does that mean for patient care quality? Learn about the controversial history of 340B, how it has shaped the American healthcare system, and the legal challenges it faces.
What happens when access to healthcare is decided by the luck of the draw? That’s exactly what Oregon tested in 2008 with a unique Medicaid selection process, revealing surprising insights about healthcare access, financial security, and patient outcomes. Learn more about this experiment’s lasting impact on health policy.